machine learning in supply chain management a systematic literature review

To read this content please select one of the options below:

Please note you do not have access to teaching notes, a systematic review of machine learning in logistics and supply chain management: current trends and future directions.

Benchmarking: An International Journal

ISSN : 1463-5771

Article publication date: 26 March 2021

Issue publication date: 5 November 2021

This paper presents a review of the existing state-of-the-art literature on machine learning (ML) in logistics and supply chain management (LSCM) by analyzing the current literature, contemporary concepts, data and gaps and suggesting potential topics for future research.

Design/methodology/approach

A systematic/structured literature review in the subject discipline and a bibliometric analysis were organized. Information regarding industry involvement, geographic location, research design and methods, data analysis techniques, university, affiliation, publishers, authors, year of publications is documented. A wide collection of eight databases from 1994 to 2019 were explored using the keywords “Machine Learning” and “Logistics“, “Transportation” and “Supply Chain” in the title and/or abstract. A total of 110 articles were found, and information on a chain of variables was gathered.

Over the last few decades, the application of emerging technologies has attracted significant interest all around the world. Analysis of the collected data shows that only nine literature reviews have been published in this area. Further, key findings show that 53.8 per cent of publications were closely clustered on transportation and manufacturing industries and 54.7 per cent were centred on mathematical models and simulations . Neural network is applied in 22 papers as their exclusive algorithms. Finally, the main focuses of the current literature are on prediction and optimization , where detection is contributed by only seven articles.

Research limitations/implications

This review is limited to examining only academic sources available from Scopus, Elsevier, Web of Science, Emerald, JSTOR, SAGE, Springer, Taylor and Francis and Wiley which contain the words “Machine Learning” and “Logistics“, “Transportation” and “Supply Chain” in the title and/or abstract.

Originality/value

This paper provides a systematic insight into research trends in ML in both logistics and the supply chain.

• Structured literature review
• Artificial intelligence
• Machine learning
• Supply chain management
• Transportation
• Emerging technologies

Akbari, M. and Do, T.N.A. (2021), "A systematic review of machine learning in logistics and supply chain management: current trends and future directions", Benchmarking: An International Journal , Vol. 28 No. 10, pp. 2977-3005. https://doi.org/10.1108/BIJ-10-2020-0514

Emerald Publishing Limited

Copyright © 2021, Emerald Publishing Limited

Related articles

We’re listening — tell us what you think, something didn’t work….

Report bugs here

All feedback is valuable

Please share your general feedback

Join us on our journey

Platform update page.

Visit emeraldpublishing.com/platformupdate to discover the latest news and updates

Questions & More Information

Answers to the most commonly asked questions here

A systematic review of machine learning in logistics and supply chain management: current trends and future directions

PurposeThis paper presents a review of the existing state-of-the-art literature on machine learning (ML) in logistics and supply chain management (LSCM) by analyzing the current literature, contemporary concepts, data and gaps and suggesting potential topics for future research.Design/methodology/approachA systematic/structured literature review in the subject discipline and a bibliometric analysis were organized. Information regarding industry involvement, geographic location, research design and methods, data analysis techniques, university, affiliation, publishers, authors, year of publications is documented. A wide collection of eight databases from 1994 to 2019 were explored using the keywords “Machine Learning” and “Logistics“, “Transportation” and “Supply Chain” in the title and/or abstract. A total of 110 articles were found, and information on a chain of variables was gathered.FindingsOver the last few decades, the application of emerging technologies has attracted significant interest all around the world. Analysis of the collected data shows that only nine literature reviews have been published in this area. Further, key findings show that 53.8 per cent of publications were closely clustered on transportation and manufacturing industries and 54.7 per cent were centred on mathematical models and simulations. Neural network is applied in 22 papers as their exclusive algorithms. Finally, the main focuses of the current literature are on prediction and optimization, where detection is contributed by only seven articles.Research limitations/implicationsThis review is limited to examining only academic sources available from Scopus, Elsevier, Web of Science, Emerald, JSTOR, SAGE, Springer, Taylor and Francis and Wiley which contain the words “Machine Learning” and “Logistics“, “Transportation” and “Supply Chain” in the title and/or abstract.Originality/valueThis paper provides a systematic insight into research trends in ML in both logistics and the supply chain.

• Related Documents

Sustainable supply chain management: continuing evolution and future directions

Purpose The purpose of this paper is to update the work of Carter and Easton (2011), by conducting a systematic review of the sustainable supply chain management (SSCM) literature in the primary logistics and supply chain management journals, during the 2010–2018 timeframe. Design/methodology/approach The authors use a systematic literature review (SLR) methodology which follows the methodology employed by Carter and Easton (2011). An evaluation of this methodology, using the Modified AMSTAR criteria, demonstrates a high level of empirical validity. Findings The field of SSCM continues to evolve with changes in substantive focus, theoretical lenses, unit of analysis, methodology and type of analysis. However, there are still abundant future research opportunities, including investigating under-researched topics such as diversity and human rights/working conditions, employing the group as the unit of analysis and better addressing empirical validity and social desirability bias. Research limitations/implications The findings result in prescriptions and a broad agenda to guide future research in the SSCM arena. The final section of the paper provides additional avenues for future research surrounding theory development and decision making. Originality/value This SLR provides a rigorous, methodologically valid review of the continuing evolution of empirical SSCM research over a 28-year time period.

Healthcare supply chain management: literature review and some issues

PurposeThe purpose of this paper is to examine the causal linkages among supply chain practices, effectiveness of supply chain performance (SCP) and organizational performance (OP) in Indian healthcare industries.Design/methodology/approachThis paper is helpful in developing a framework for linking a healthcare supply chain practice to its OP, and thus identifies how such a linkage can be connected to the effectiveness of SCP. Such effort also enables the authors to derive a set of recommended supply chain practices for SC performance.FindingsFrom the literature review, this paper finds that, in the context of Indian healthcare industries, efficient SC performance may play a critical role for overall OP improvement, as there is a close interrelationship between supply chain management (SCM) practices and SCP that may have a more significant effect on OP improvement.Research limitations/implicationsThe principle limitation of the paper is that it is performed only in a particular industry and with a questionnaire survey which could be extended in future for other industries also. Another limitation of the paper is that it is focused only on the SCP of medical device and equipment supply chain which is a small portion of the whole healthcare supply chain, and therefore requires further research covering various other domains of healthcare supply chain. Another limitation of the study is that the sample survey has been taken from only one respondent per company at one point of time which may create biasness in the results. Thus, future research should collect data through multiple members from the organization.Practical implicationsThis study contributes to know the effect of SCM practices on healthcare SCP and provides a practical and useful tool to evaluate the extent of effectiveness of SCP and finally their impact on the healthcare OP. Finally, this study provides conceptual and descriptive literature regarding SCM practices that leads to improvement in healthcare performance.Social implicationsThis study adds to the knowledge on healthcare SCM performance by exploring the relationship between supply chain practices, healthcare SCP and healthcare OP and by developing and testing a research framework thus help in improving patient satisfaction.Originality/valueThis study attempts to show how the potential benefits of supply chain practices can no longer be ignored in healthcare supply chain.

An analysis of barriers affecting implementation of sustainable supply chain management in electronics industry: a Grey-DEMATEL approach

Purpose World over organizations are focusing on sustainable goals, where along with economic success their role in protecting the planet and people are becoming important. Whilst transforming the supply chain into a sustainable one, there would be some barriers which might hinder this process. This paper aims to study these barriers in the context of the electronics industry so that organizations can better implement sustainable supply chain programs. Design/methodology/approach In this research, barriers affecting sustainability implementation in the electronics supply chain are shortlisted from literature review and experts’ opinion. Using the combined methodology of Grey DEMATEL, the causal factors, the effect factors and degree of prominence of barriers is found out. The overall relationship among barriers is established by a diagraph. Sensitivity analysis is performed to check the robustness of the results. Findings It is found that lack of regulation and guidance from authorities is the primary causal barrier affecting operations of sustainable supply chain management. There are five barriers which fall in the influenced group and among them, complexity in measuring and monitoring sustainability practices has the largest net effect value on the implementation of a sustainable supply chain. The barrier having the highest correlation with other barriers is the high cost for disposal of hazardous wastes. The implications of these findings on managers and academicians is explored in the study. Research limitations/implications In this research, the number of barriers shortlisted is limited to 11 in the context of the electronics supply chain. More factors could be added in future research based on the industry being studied. Originality/value The research analyses 11 barriers under categories of policy, technology, financial and human resources in the Indian electronics industry by evaluating the cause and effect group of barriers. These results can guide policymakers of the electronic sector and industry for mitigating barriers during the implementation of sustainable programs.

Contrasting supply chain traceability and supply chain visibility: are they interchangeable?

PurposeSupply chain traceability and supply chain visibility have become a critical element for the effective management of contemporary complex supply chains. At their core is information sharing, which has been acknowledged as a key prerequisite for logistics and supply chain performance, but whose notional underpinnings have not been delineated fully, leading to interchangeable deployment of these terms. Addressing the shortcoming, this paper aims to establish a contrast between the two notions.Design/methodology/approachDrawing from systematic review protocols, a multi-disciplinary review scope is constructed wherein the synthesis is strategized to primarily channel implications for the scholarship of logistics and supply chain management. The review is aimed at addressing two research objectives: (1) how the notions of traceability and visibility in supply chain management develop contrast in terms of their thematic emphasis and (2) to attain an integrative understanding of the notional convergence and divergence between supply chain traceability and visibility for raising strategic recommendations.FindingsThe review outcomes help contrast both the convergence and the divergence between traceability and visibility in the supply chain environment, and the differentiated but fundamental role that information sharing plays within these notions to outline why they are not interchangeable.Originality/valueThe originality of the findings lies in the conceptual synthesis of the relevant literature from both technological and non-technological perspectives to ultimately draw logistics and supply chain management implications. The review also points out key strategic considerations to demarcate the notional boundaries of traceability and visibility in future research.

Internal integration in humanitarian supply chain management

Purpose The purpose of this paper is to compare perspectives on humanitarian logistics (HL) and supply chain management (SCM) among programmes and logistics/support staff. Design/methodology/approach Underpinned by services supply chain management (SSCM) theory, a single case study of a leading international non-governmental organisation is presented based on a web-based survey of the organisation’s global operations staff, supplemented by semi-structured interviews conducted with senior representatives. Findings The study is believed to be the first to consider the different perspectives of programmes and logistics staff on the interpretation of logistics and SCM. The results indicate both significant divergence between the views of these two cohorts, as well as a general lack of clarity over the concept of SCM, its relationship with logistics and the cross-functional nature of SCM. Research limitations/implications Insufficient responses from programme staff limit the generalisability of the findings. Suggestions for future research include further examination of the potential of applying SSCM and demand chain management concepts to the humanitarian context. Practical implications The results support the notion that a broader, more strategic interpretation of SCM, more clearly distinguished from the practice of HL, may assist in breaking down perceived jurisdictional boundaries, bridging the gap between programmes and logistics teams, and strengthening demand-chain influences and the “voice of the beneficiary”. Originality/value By taking into account the views of non-logisticians, a broader, cross-functional interpretation of SCM is offered leading to revised definitions for both SCM and HL within this sector, together with a framework that integrates SCM across humanitarian relief and development contexts.

Lean resilience: AURA (Active Usage of Resilience Assets) framework for post-COVID-19 supply chain management

PurposeSupply chain resilience capabilities are usually considered in light of some anticipated events and are as passive assets, which are “waiting” for use in case of an emergency. This, however, can be inefficient. Moreover, the current COVID-19 pandemic has revealed difficulties in the timely deployments of resilience assets and their utilization for value creation. We present a framework that consolidates different angles of efficient resilience and renders utilization of resilience capabilities for creation of value.Design/methodology/approachWe conceptualise the design of the AURA (Active Usage of Resilience Assets) framework for post-COVID-19 supply chain management through collating the extant literature on value creation-oriented resilience and practical examples and complementing our analysis with a discussion of practical implementations.FindingsBuilding upon and integrating the existing frameworks of VSC (Viable Supply Chain), RSC (Reconfigurable Supply Chain) and LCNSC (Low-Certainty-Need Supply Chain), we elaborate on a new idea in the AURA approach – to consider resilience as an inherent, active and value-creating component of operations management decisions, rather than as a passive “shield” to protect against rare, severe events. We identify 10 future research areas for lean resilience integrating management and digital platforms and technology.Practical implicationsThe outcomes of our study can be used by supply chain and operations managers to improve the efficiency and effectiveness by turning resilience from passive, cost-driving assets into a value-creating, inclusive decision-making paradigm.Originality/valueWe propose a novel approach to bring more dynamics to the notion of supply chain resilience. We name our approach AURA and articulate its two major advantages as follows: (1) reduction of disruption prediction efforts and (2) value creation from resilience assets. We offer a discussion on ten future research directions towards a lean resilience.

Understanding supply chain analytics capabilities and agility for data-rich environments

Purpose Big data-driven supply chain analytics capability (SCAC) is now emerging as the next frontier of supply chain transformation. Yet, very few studies have been directed to identify its dimensions, subdimensions and model their holistic impact on supply chain agility (SCAG) and firm performance (FPER). Therefore, to fill this gap, the purpose of this paper is to develop and validate a dynamic SCAC model and assess both its direct and indirect impact on FPER using analytics-driven SCAG as a mediator. Design/methodology/approach The study draws on the emerging literature on big data, the resource-based view and the dynamic capability theory to develop a multi-dimensional, hierarchical SCAC model. Then, the model is tested using data collected from supply chain analytics professionals, managers and mid-level manager in the USA. The study uses the partial least squares-based structural equation modeling to prove the research model. Findings The findings of the study identify supply chain management (i.e. planning, investment, coordination and control), supply chain technology (i.e. connectivity, compatibility and modularity) and supply chain talent (i.e. technology management knowledge, technical knowledge, relational knowledge and business knowledge) as the significant antecedents of a dynamic SCAC model. The study also identifies analytics-driven SCAG as the significant mediator between overall SCAC and FPER. Based on these key findings, the paper discusses their implications for theory, methods and practice. Finally, limitations and future research directions are presented. Originality/value The study fills an important gap in supply chain management research by estimating the significance of various dimensions and subdimensions of a dynamic SCAC model and their overall effects on SCAG and FPER.

An ISM approach for the barrier analysis in implementing sustainable supply chain management

Purpose Sustainability has been on the executive agenda for years and it is now one of the fastest growing supply chain management trends. The purpose of this paper is to analyze the barriers for the adoption and implementation of the sustainable supply chain management (SSCM) concept. Design/methodology/approach This study has been divided into two phases such as identification of barriers and qualitative analysis. First, to identify the most influential barriers, the authors offer a systematic literature review, taking 188 papers published from 2010 to November 2016 into account. The investigation phase led to the selection of 15 barriers based on the literature in consultation with industrial experts and academicians. Second, the interpretive structural modeling qualitative analysis was used to find out the mutual influences between the 15 barriers by a survey. Findings Further, the authors propose and illustrate the cross-impact matrix multiplication applied to classification analysis to test a framework that extrapolates SSCM barriers and their relationships. “Inadequate information technology implementation” has been identified as the most important barrier that may force organizations to implement SSCM practices to ensure their business sustainability. Research limitations/implications The authors presented some limitations in their research in some fields which could allow new researchers and practitioners to conduct the future research to grow in different dimensions. Practical implications Practitioners or policymakers usually are not familiar with these types of research works; that is why most of these surveys remain theoretical and conceptual. Future investigation needs to be done in practical application domain instead of merely giving opinions. Originality/value Based on the authors’ research, the researchers have more attention to work in conceptual analysis due to other fields, but the authors believe that even with the implementation of SSCM, many remarkable areas still exist for future research which could help in development. The authors also provide more details in this paper.

Dynamic capabilities in humanitarian supply chain management: a systematic literature review

PurposeThe main purpose of this paper is to examine the extant literature of humanitarian supply chain management (HSCM) which specifically use dynamic capabilities (DCs) view. By this means, the objectives of this study are to identify and assess the DCs used in the HSCM context, the factors positively and negatively affecting the DCs and how the DCs affect humanitarian supply chain (HSC) operations. Furthermore, this research aims to give directions for future research in the field of HSCM.Design/methodology/approachThis study adopts systematic literature review (SLR) approach proposed by Denyer and Tranfield (2009). Based on a SLR, this study synthesizes and compares the evidence, has a specific focus and research questions, has certain inclusion and exclusion criteria and provides evidence-based implications to the researchers and practitioners. This is a method which is replicable, transparent and auditable. The SLR methodology provides scholars and practitioners a basis for comprehending the current situation of relevant topic and taking correct steps in their future actions.FindingsThis SLR deduces that applying DCs view is still in its infancy in the HSCM literature. The result of this SLR reveals that supply chain agility (SCA), supply chain resilience (SCR), reconfiguration/transformation, integration, (short-term) collaboration, sustaining, sensing, seizing and knowledge access DCs have been used in the HSCM literature. In addition, it is determined that only one paper analyzed the influence of DCs on predisaster performance while rest of the papers focused on the postdisaster performance.Originality/valueThe result of the exhaustive literature search indicates that this is the first SLR that specifically analyzes the application of DCs view in the HSCM domain. This investigation determined the DCs used in HSCM and revealed the relations between the dependent and independent variables through the comprehensive model. In this way, this review provides a guidance to researchers in conduct their future investigations and practitioners to carry out supply chain (SC) operations by considering the factors affecting their operations.

Doctoral dissertations in supply chain management: a review of Chinese contributions from 1995 to 2019

PurposeThis study aims to investigate Chinese doctoral dissertations in supply chain management (SCM) by providing an overview of the evolution of this topic over the past 25 years.Design/methodology/approachThis study reviews 1,945 doctoral dissertations in SCM from 39 Project 985 universities in China.FindingsThe annual number of SCM doctoral dissertations in China varied significantly over time, and the top producers are all strong players in the SCM field with long histories of research. The dissertations use multiple methods, of which mathematical, econometric modeling and simulation, and survey are the most frequently used. Most of the frequently used theories are borrowed from other disciplines. Traditional research topics are frequently investigated in Chinese doctoral dissertations in SCM, but the popular research topics differ by period.Originality/valueThis study has examined Chinese doctoral dissertations in SCM through a literature review. This study provides implications for research topics, methodological approaches, and theoretical lenses for future research. It also disseminates Chinese SCM knowledge to the global academic network to stimulate more SCM research.

Export Citation Format

Share document.

HTML conversions sometimes display errors due to content that did not convert correctly from the source. This paper uses the following packages that are not yet supported by the HTML conversion tool. Feedback on these issues are not necessary; they are known and are being worked on.

• failed: apalike

Authors: achieve the best HTML results from your LaTeX submissions by following these best practices .

AI in Supply Chain Risk Assessment: A Systematic Literature Review and Bibliometric Analysis

Supply chain risk assessment (SCRA) has witnessed a profound evolution through the integration of artificial intelligence (AI) and machine learning (ML) techniques, revolutionizing predictive capabilities and risk mitigation strategies. The significance of this evolution stems from the critical role of robust risk management strategies in ensuring operational resilience and continuity within modern supply chains. Previous reviews have outlined established methodologies but have overlooked emerging AI/ML techniques, leaving a notable research gap in understanding their practical implications within SCRA. This paper conducts a systematic literature review combined with a comprehensive bibliometric analysis. We meticulously examined 1,717 papers and derived key insights from a select group of 48 articles published between 2014 and 2023. The review fills this research gap by addressing pivotal research questions and exploring existing AI/ML techniques, methodologies, findings, and future trajectories, thereby providing a more encompassing view of the evolving landscape of SCRA. Our study unveils the transformative impact of AI/ML models, such as Random Forest, XGBoost, and hybrids, in substantially enhancing precision within SCRA. It underscores adaptable post-COVID strategies, advocating for resilient contingency plans and aligning with evolving risk landscapes. Significantly, this review surpasses previous examinations by accentuating emerging AI/ML techniques and their practical implications within SCRA. Furthermore, it highlights the contributions through a comprehensive bibliometric analysis, revealing publication trends, influential authors, and highly cited articles. This analysis provides an encompassing view of the field’s growth, offering insights into its evolution. This comprehensive review provides a roadmap for practitioners and researchers, offering insights into fortifying supply chain risk management strategies through AI integration, ultimately contributing to a deeper understanding of evolving trends and applications in this dynamic field.

[inst1]organization=Department of Industrial Engineering and Management,addressline=Khulna University of Engineering and Technology (KUET), city=Khulna, postcode=9203, country=Bangladesh

[inst2]organization=Department of Computer Science,addressline=American International University-Bangladesh (AIUB), city=Dhaka, postcode=1229, country=Bangladesh

1 Introduction

Supply chain management (SCM) has long grappled with various challenges stemming from a spectrum of sources. Past outbreaks of infectious diseases, geological upheavals like earthquakes, and other natural catastrophic events have put supply chains at risk, albeit on a limited scale (Govindan et al.,, 2020 ) . These incidents have illustrated how interrelationships within supply chains have formed intricate risk contagion networks, making them susceptible to contagion effects (Agca et al.,, 2022 ) . The cascade effect, characterized by risk spillover from one enterprise to another, exacerbates the challenges, culminating in a chain reaction of Supply Chain Resilience (SCR) disasters (Roukny et al.,, 2018 ) .

Across the globe, companies confront formidable challenges across all stages of their supply chains. Suppliers failing to meet delivery obligations, unpredictable shifts in customer demands, and episodes of panic buying are just a few examples of the hurdles companies face (Ivanov,, 2020 ) . In pursuing market leadership, enterprises must also contend with the complexity introduced by the management and launch of innovation projects, further complicating SCM (Kwak et al.,, 2018 ) . Furthermore, digital transformation has ushered in a wave of technological advancements in supply chain operations (Kwak et al.,, 2018 ) .

However, the most significant disruption in recent times emerged during the COVID-19 pandemic, impacting global supply chains profoundly. The disruptions encompassed not only the movement of people but also the flow of raw materials and finished goods, along with extensive disruptions in factory and supply chain operations (Sheng et al.,, 2021 ) . These disruptions ushered in unprecedented challenges for supply chain professionals as they grappled with an entirely new reality (Araz et al.,, 2020 ; Craighead et al.,, 2020 ) . The pandemic represents a distinctive form of supply chain disruption, distinct from natural or man-made disasters (e.g., the Japanese earthquake and 9/11 attacks) and disruptions driven by evolving technologies and changing customer attitudes (Zhang et al.,, 2020 ) .

While existing guidance exists for predicting, managing, and responding to these disruptions, the challenges brought about by COVID-19 have underscored the paramount importance of risk management, unlike any previous disruption (Bode et al.,, 2011 ; Craighead et al.,, 2020 ) . Consequently, the concept of SCR in the context of COVID-19 has emerged as a critical area necessitating further exploration and development.

Supply chains are not the sole victims in this precarious environment, as the ripple effect extends to upstream and downstream enterprises. Practical risk management programs are imperative for enterprises to avert SCR and enhance SCR management (SCRM). Moreover, incorporating advanced AI technologies, such as ML, can provide predictive capabilities for SCRM. ML algorithms have showcased their ability to identify abnormal risk factors and derive predictive insights from historical data (Guo et al.,, 2021 ; Mohanty et al.,, 2021 ) . By harnessing ML, enterprises can detect risk factors and anticipate market demands and potential risk scenarios (Punia et al.,, 2020 ; Wu et al., 2022a, ) . ML’s proficiency in handling non-linear relationships further bolsters its superiority over traditional linear models. Additionally, its aptitude for processing unstructured data, a challenge where traditional models falter, positions ML as a formidable tool for addressing time, cost, and resource constraints within supply chains.

However, despite its immense potential, supply chain researchers have historically exhibited limited familiarity with ML in comparison to other SCM aspects like mathematical programming and stochastic optimization (Liu et al.,, 2019 ; Shahed et al.,, 2021 ) . Furthermore, the classification of ML algorithms remains unclear (Janiesch et al.,, 2021 ; Xu and Jackson,, 2019 ) . Bridging this knowledge gap and exploring the value of ML for SCRM through interdisciplinary integration is a critical research need. Remarkably, there has been no prior effort to scrutinize the SCRM literature within an ML environment, in stark contrast to previous reviews primarily centered on risk definition, classification, and management strategies.

Companies should augment their analytical capabilities by harnessing organizational knowledge to enhance SCR, thereby elevating their information capabilities (Wong et al.,, 2020 ) . As underscored by previous studies, the role of AI-enabled technologies extends to promoting innovations for enhanced supply chain performance (SCP) ( Baryannis et al., 2019b, ; Nayal et al.,, 2021 ) . The adaptation capabilities and information processing prowess offered by AI techniques hold the potential to enhance SCP ( Belhadi et al., 2021a, ) . Notably, AI has found application across diverse sectors, improving flexibility and communication and reducing undue fluctuations for successful project execution (Lalmi et al.,, 2021 ) . To navigate the impact of risks and disruptions, (Katsaliaki et al.,, 2022 ) advocate the integration of three critical facets: long-term partnerships, IT applications for business enhancement, and government policies that facilitate adaptability.

The contemporary supply chain landscape demands the integration of AI, specifically ML, to invigorate SCRA and SCRM. While traditional methods possess merit, their limitations can be effectively addressed by AI, which offers predictive capabilities, nonlinear relationship analysis, and unstructured data processing. As supply chains continue to evolve, AI-based risk assessment is imperative for fostering resilience and sustaining the efficiency and effectiveness of supply chain operations. Existing reviews of SCRA using ML often fall short of offering a comprehensive view of evolving ML techniques, neglecting emerging trends and overlooking diverse applications and non-journal sources. They also tend to overlook the role of ML in the response phases of SCRM and do not sufficiently address data-related challenges.

In light of the limitations associated with traditional SCRA methods, there is a growing interest in harnessing the potential of AI techniques to enhance risk assessment practices. This paper presents a systematic literature review (SLR) focused on the application of AI in SCRA. The main objective of this review is to analyze the existing literature critically, identify research gaps, and provide insights into the use of AI techniques, such as ML, deep learning(DL), and natural language processing(NLP), to improve the accuracy and effectiveness of SCRA. Our paper has the following contributions-

Our study significantly contributes to the literature by conducting a comprehensive review and employing bibliometric and cocitation analysis to assess the application of AI and ML techniques in SCRA, specifically focusing on the most recent papers.

This research addresses the evolving landscape of SCRA by considering the challenges posed by post-COVID uncertainties and emphasizing the increasing significance of AI in risk assessment.

This systematic review paper addresses the evolving landscape of SCRA by considering the challenges posed by post-COVID uncertainties and emphasizing the increasing significance of AI in risk assessment.

Our paper highlights pivotal trends and key findings, shedding light on the current landscape of AI/ML in SCRA.

We offer valuable insights into the next steps for researchers and practitioners, guiding the evolution of SCRA methods in an ever-changing world.

Our review aids academics and industry professionals seeking effective risk management strategies in today’s dynamic supply chain environment.

This review follows a structured paper skeleton, starting with an introduction to the topic and its significance in the field of SCM in the “ Introduction ” section. Following this, a comprehensive “ Literature Review ” unfolds, delving into the existing body of knowledge. The subsequent “ Methodology ” section outlines the study’s design, including article search strategies, selection criteria, and the synthesis of gathered data. Complementing these fundamental segments, the paper integrates a specific section for “ Bibliometric Analysis ” Utilizing bibliometric tools, this section showcases insights distilled from a comprehensive analysis of scholarly publications, focusing on AI in SCRA. Visualizations and key observations derived from this analysis are encapsulated here. Transitioning forward, the narrative navigates through diverse AI techniques deployed in SCRA, expounded within the “ AI Techniques for Supply Chain Risk Assessment ” section. Here, the discussion encapsulates the advantages, limitations, and real-world applications of these techniques. Moreover, it underlines the “ Managerial Implications ”, providing actionable insights and recommendations for industry practitioners based on the reviewed literature. This section delineates practical implications derived from the reviewed scholarly material aimed at aiding managerial decision-making within SCRA contexts. Continuing, the paper addresses the “ Challenges and Limitations ”encountered in this field, and forecasts potential “ Future Research Directions ” This forward-looking section paves the way for continued exploration and advancement within this domain. Ultimately, the review culminates in a comprehensive “ Conclusion ”, summarizing the primary findings drawn from the review process. Additionally, it furnishes recommendations tailored for both practitioners and researchers aimed at leveraging the full potential of AI in SCRA.

This meticulously structured approach endeavors to contribute to existing knowledge and bridge critical gaps within the literature. It is designed to furnish invaluable insights for academia and industry stakeholders alike.

2 Literature Review

Deiva Ganesh and Kalpana, ( 2022 ) conducted a comprehensive and descriptive literature study to identify AI and ML approaches in SCRM stages. This study examined SCRM research publications from three scientific databases from 2010 to 2021. They proposed a data analytics, simulation, and optimization framework that might produce a comprehensive Supply Chain risk identification, assessment, mitigation, and monitoring strategy. Hence, leveraging AI, Blockchain, and the Industrial Internet of Things (IIoT) to create smarter supply chains can transform how firms handle uncertainty. This analysis of (Deiva Ganesh and Kalpana,, 2022 ) does not evaluate blockchain, big data, and IIoT-related supply chain articles, which may limit information. Due to limited research, only recent English-language publications on SCRM and AI were included. Nimmy et al., ( 2022 ) did a thorough literature evaluation on operational risk assessment methods and whether AI explains them. The report suggested that risk managers use auditable supply chain operational risk management methodologies to understand why they should take a risk management action rather than just what to do. While they employed the SLR method, they could only examine some AI solutions for SCRM. They suggested hybridizing supply chain operational risk management and explainable AI (XAI) to obtain XAI-like characteristics in future research. Li et al., ( 2023 ) analyzed COVID-19 SCR research history, present, and future. In particular, supervised ML classifies 1717 SCR papers into 11 subject categories. Each cluster was then studied in the context of COVID-19, indicating three related skills (interconnectedness, transformability, and sharing) on which enterprises could work to develop a more resilient supply chain post-COVID. Their data only came from Scopus’ core collection, which might affect the results; thus, they suggested adding WoS and EBSCO to the evaluation Xu et al., ( 2020 ) . Given the fast expansion of SCR research, they only picked English-language articles, which may have excluded valuable knowledge. They recommended network analysis to determine cluster linkages and SCR literary themes.

Naz et al., ( 2022 ) examined the role of AI in building a robust and sustainable supply chain and offered optimal risk mitigation options. For review, 162 SCOPUS research publications were selected. Based on the nominated articles, Structural Topic Modeling was used to produce various AI-related theme topics in SCR. AI research trends in SCR were examined using R-package bibliometric analysis. They solely studied journal articles, not conference papers, field reports, corporate reports, book chapters, etc. Ni et al., ( 2020 ) analyzed articles from 1998/01/01 to 2018/12/31 in five major databases to highlight the newest research trends in the field. ML applications in SCM were still developing due to a lack of high-yielding authors and poor publication rates. 10 ML algorithms were extensively utilized in SCM, but their utilization was unequal among the SCM tasks most often reported. This paper has limitations in reviewing only five popular databases to limit articles for evaluation that might have filtered some related articles. Second, only widely used ML methods in SCM were counted in this review, and other ML techniques brought to SCM may be helpful later on. Low-frequency ML algorithms should be analyzed for further research in this area. Finally, their article contained 32 well-recognized ML methods; however, some newly generated ML algorithms may be used in SCM after 2015. Baryannis et al., 2019b examined supply chain risk definitions and uncertainty. Then, a mapping analysis categorizes available literature by AI approaches and SCRM tasks. Most of the examined works focus on building and assessing a mathematical model that accounts for various uncertainties and hazards but less on establishing and analyzing the applicability of the suggested models. They found that only 9 of 276 research (3%) use comprehensive techniques that cover all three SCRM stages (identification, assessment, and response). Yang et al., ( 2023 ) thoroughly examined the advancement of ML algorithms in SCRM by gathering 67 publications from 9 authoritative databases in the first half of 2021. They analyzed only English language journal articles from 9 relevant academic databases, excluding conference papers, textbooks, and unpublished articles and notes. This analysis relied on keyword searches, which may have missed some work. Schroeder and Lodemann, ( 2021 ) used a comprehensive and multi-vocal literature study to get a complete picture of our subject field. The SLR identified 533 papers in this area and examined 23. The comprehensive literature study found that just a few examples of ML in SCRM have been reported in depth in the scientific literature, and those examples focus on manufacturing, transport, and the complete supply chain. Shi et al., ( 2022 ) examined 76 works on credit risk utilizing statistical, ML, and DL methods over the last eight years. They offered a unique classification approach and performance rating for ML-driven credit risk algorithms utilizing public datasets. Data imbalance, dataset inconsistency, model transparency, and DL model underuse are discussed. Their review found that most DL models outperform standard ML and statistical algorithms in credit risk prediction, and ensemble techniques outperform single models.

Research Questions We formulated the following research questions to guide our SLR:

What is the current state of research on applying AI/ML techniques in SCRM?

Which AI/ML techniques are commonly employed in supply chain risk assessment, prediction, and mitigation?

What are the key findings and trends identified in the literature regarding using AI/ML in SCRM?

What are the research gaps and future directions in this field?

3 Methodology

3.1 study design.

The research utilized a systematic approach to conduct a comprehensive literature review coupled with a bibliometric analysis. This methodology aimed to scrutinize the intersection of AI/ML techniques and SCRA, providing insights into research trends, methodologies, and emerging themes.

The SLR followed a robust methodology, adhering to the guidelines proposed by (Denyer and Tranfield,, 2009 ) . This approach aimed to identify and analyze research articles focusing on integrating AI/ML techniques within SCRA.

3.2 Search Strategy

We conducted a comprehensive search using two major academic databases: Google Scholar and Web of Science (WoS). We used the ”Publish or Perish 8” tool for the Google Scholar search to filter articles published between 2015 and 2023. The initial search yielded 1000 articles. In the WoS search, we employed the following search strings to retrieve relevant articles:

(ALL=((”supply chain”) AND (”risk*” OR ”credit risk*” OR ”risk assessment*” OR ”risk prediction” OR ”disruption*” OR ”disease outbreak*” OR ”post COVID resilienc*”) AND (”artificial intelligence” OR ”machine learning” OR ”neural network*” OR ”deep learning” OR ”reinforcement learning” OR ”SVM” OR ”support vector machine” OR ”boosting” OR ”ensemble” OR ”bayesian network model” OR ”random forest” OR ”LSTM” OR ”long short-term memory”))) NOT (DT==(”PROCEEDINGS PAPER” OR ”BOOK CHAPTER” OR ”EDITORIAL MATERIAL” OR ”LETTER”))

The WoS search initially retrieved 434 articles.

3.3 Article Selection Process

We followed a systematic approach to select the most relevant articles for our analysis. The inclusion and exclusion criteria were applied sequentially to ensure the selection of high-quality research. The following steps were taken:

Removal of Duplicates: We removed duplicate articles identified between the two databases, resulting in 360 duplicates.

Filtering by Reliable Publishers: We only considered articles published by reliable and reputable publishers, excluding 398 articles.

Non-English Exclusion: We excluded articles not written in English, resulting in removing 3 articles.

Filtering by Publication Type: We excluded articles classified as Q3 or Q4, proceedings papers, book chapters, editorials, and letters, resulting in the removal of 244 articles.

Title, Abstract, and Conclusion Review: We reviewed the remaining articles based on their title, abstract, methodology, and conclusion to ensure their relevance to our research questions.

This step removed 386 articles, leaving us with a final set of 48 research articles.

3.4 Data Extraction and Synthesis

The selected articles underwent detailed analysis, focusing on methodologies, AI/ML techniques employed, key findings, and implications for SCRA. Data extraction forms were filled in to gather pertinent information for synthesis and further analysis.

To understand the research landscape comprehensively, we performed bibliometric and cocitation analysis of the 434 articles resulting from the filtering process. The bibliometric analysis helped us identify the field’s most influential authors, journals, and key research themes. Cocitation analysis allowed us to identify clusters of related articles and determine their interconnections.

Our study has several limitations that should be acknowledged. First, the search was limited to articles published between 2015 and 2023, and it is possible that relevant articles published before or after this timeframe were not included.

4 Bibliometric Analysis

In our quest to comprehensively analyze the landscape of supply chain risk assessment using machine learning, we employed an array of tools, including MS Excel and VOSviewer software. This bibliometric analysis comprised several key components, each contributing to a deeper understanding of the field.

4.1 Publication Trend Over Time

Figure  2 offers a dynamic portrayal of the publication trend over the years, presenting an invaluable temporal dimension to our bibliometric analysis. A meticulous examination of data showed how the research landscape has evolved from 2014 to 2023. This longitudinal perspective equips us with the knowledge of emerging trends, shifts in focus, and the growth trajectory of the field.

4.2 Geographical Distribution of Research

In Figure  3 , we present a geographical distribution of research related to SCRA using machine learning. The figure offers a clear visualization of countries that have actively contributed to the literature in this field. The deeper coloration signifies a higher volume of research publications from these regions. This representation enriches our understanding of the global landscape of SCRA, highlighting regions with significant scholarly contributions.

4.3 Analysis of Authors

Table   1 presents a ranking of the top 10 authors in the field of risk prediction and analysis based on the number of research papers attributed to each. ”Gupta S” leads the list with 10 papers, highlighting a substantial body of work in this domain. This table serves as a concise reference for identifying influential authors who have made significant contributions to the field of risk analysis and provides valuable insights into their scholarly output, aiding researchers and professionals in exploring their work.

4.4 Publisher Contributions

Figure  4 provides a compelling visual representation of the leading publishers in the field of SCRA, with a specific emphasis on applications of AI and ML. This chart clearly depicts the number of research papers associated with each publisher, enabling researchers and professionals to discern the most prolific contributors to this dynamic domain. It serves as a valuable resource for identifying key publishers that disseminate influential research at the intersection of AI/ML and SCRA, granting a panoramic view of the publication landscape in this evolving field.

4.5 Journal Contributions

This bar chart, represented in Figure  5 , showcases the top 11 journals significantly contributing to the field of SCRA with a specific focus on AI and ML applications. This visual representation helps researchers and professionals identify the key journals for accessing relevant and impactful research in AI/ML and SCRA. It concisely overviews this dynamic field’s most prolific journal sources.

4.6 Co-citation Analysis

In pursuing a comprehensive understanding of the scholarly landscape in AI-based SCRA, we present three co-citation analyses derived from bibliometric data of the authors, cited references, and publication sources using VOSviewer. Each visualization offers unique insights into collaboration, thematic prominence, and citation patterns within this dynamic field.

Figure  5(a) illustrates connections between authors, highlighting 55 out of 18,858 authors surpassing a minimum citation threshold of 20. These authors formed three distinct clusters based on their co-citation patterns: Red (22 items), Green (19 items), and Blue (14 items). The clusters signify groups of authors who are frequently cited together, suggesting shared research themes or collaboration networks. Node size signifies citation counts, and connections depict co-citation relationships. This visualization unveils collaborative networks and shared contributions among influential authors.

The co-citation analysis of cited references involved a minimum citation threshold of 10, identifying 58 cited references meeting this criterion (Figure 5(b) ). These references formed five distinct clusters, reflecting thematic relationships among the cited works. The clusters shed light on the interconnectedness of references, providing insights into key themes or influential works within the field. Five formed clusters and individual nodes within each analysis were carefully examined for meaningful patterns. Each cluster represents distinct research themes or methodologies, while individual nodes signify specific authors, countries, or references that play pivotal roles in the overall co-citation network.

A minimum of 50 citations per source was set to analyze publication sources, identifying 70 sources meeting the criteria (Figure 5(c) ). These sources were grouped into three clusters based on their co-citation patterns. These clusters offer insights into the thematic literature concentrations within your field of study. The three formed clusters were examined to discern the major themes and sub-fields represented by the co-cited sources. The significance of central sources was considered within each cluster, as they represented foundational works or widely acknowledged references in the field.

4.7 Co-authorship Analysis

Co-authorship analysis is a method to explore and understand collaborations among authors or researchers based on their joint publication activities. This analysis highlights the collaborative essence inherent in this research domain, revealing influential partnerships among authors and the noteworthy contributions of specific countries.

Figure 6(a) unveils the co-authorship network among authors within this field. By setting a threshold of at least 3 documents per author, we identified 63 authors meeting this criterion. The resulting visualization, created using VOSViewer, showcased 6 distinct colored clusters. For instance, clusters colored red, green, and deep blue comprised 10, 7, and 5 authors, respectively. The clusters are formed based on similar collaborative behaviors among these authors. The links connecting authors within clusters represent the robustness of their collaboration, indicating frequent joint authorship or shared research endeavors.

Figure 6(b) demonstrates co-authorship analysis at a country level. Using data encompassing document count, citations, and total link strength, a minimum document threshold of 2 per country identified 47 countries meeting this criterion. The visualization prominently emphasizes China’s substantial dominance across document count, citations, and total link strength compared to other countries. This dominance underscores China’s extensive contributions to the realm of Supply Chain Risk Assessment utilizing AI/ML techniques.

4.8 Co-occurrence Analysis

The results of our co-occurrence analysis using VOSviewer, focusing on the unit of analysis as ’Authors’ and ’Keywords Plus,’ were presented in Figure  8 , offering insights into the thematic intricacies of SCRA and ML. The co-occurrence analysis is a powerful tool to reveal latent connections and thematic clusters within the expansive landscape of scholarly literature. VOSviewer facilitated the exploration of how certain keywords converge within the existing body of literature. This visualization enables us to discern the prevalent themes and areas of emphasis within this research space, guiding our understanding of the critical topics that have garnered scholarly attention.

In Figure 7(a) , the network map illustrates the relationships among these keywords based on their co-occurrence within the authorship data. Each node represents a keyword, and the links between nodes signify the strength of co-occurrence. Larger nodes indicate keywords with higher occurrences, while the proximity of nodes reflects the degree of association. In this analysis, 1483 keywords were initially considered, but after applying a minimum threshold of 4 occurrences, 44 keywords emerged as frequent keywords.

In Figure 7(b) , the analysis was conducted by setting a minimum threshold of 10 occurrences for each keyword, identifying 42 keywords that surpassed this criterion from a total pool of 1032 keywords. The 42 keywords meeting the established threshold were subjected to clustering analysis, unveiling intricate co-occurrence patterns. These 4 clusters signify prevalent themes and provide insights into the interconnectivity of concepts within our research domain.

4.9 Top 10 Affiliations

Table   2 provides an overview of the top 10 affiliations based on paper count in the field of SCRA. These affiliations have significantly contributed to the literature, collectively accounting for a notable portion of the total papers reviewed (2.995% to 1.613%). The table underscores the involvement of various academic and research institutions in advancing research related to SCRA.

4.10 Bibliographic Coupling

Bibliographic coupling is a method in bibliometrics that assesses the similarity between documents based on their shared references. This approach identifies connections and relationships among scholarly works by analyzing the overlap in their citation patterns. In other words, if two documents cite a similar set of references, they are considered to be bibliographically coupled. This analysis helps reveal thematic associations, collaborative efforts among authors, and the intellectual structure of a research field. Clusters formed through bibliographic coupling indicate groups of documents, authors, countries, organizations, or sources that share common references, suggesting a degree of interrelatedness in their research content. The strength and nature of these couplings provide valuable insights into the structure and dynamics of academic knowledge networks.

4.10.1 Authors

In Figure  9 , we identified 39 significant authors out of 1407, forming six distinct clusters based on a minimum document threshold of 3 per author. These clusters represent cohesive groups of authors who share common research interests, as evidenced by the co-citation of their works. Here influential contributors like ”Gupta, Shivam,” ”Ivanov, Dmitry,” ”Modgil, Sachin,” ”Qu, Yingchi,” ”Wang, Gang-Jin,” ”Xie, Chi,” and ”Zhu, You” were identified, signifying their substantial impact within cohesive clusters of authors, emphasizing shared research interests through their co-cited works.

4.10.2 Countries

Figure  10 revealed 39 nations out of 67 that met the criteria, forming 6 clusters based on a minimum document threshold of 3 per country. These clusters provide insights into international collaborations and thematic concentrations among countries in scholarly publications. Standing out prominently are the People’s Republic of China and major contributors like England, Germany, and the USA, who demonstrate extensive involvement and collaborations in this domain, highlighting significant thematic concentrations among nations in scholarly publications.

4.10.3 Documents

In Figure  11 , 109 significant papers out of 434 met the criteria, forming 8 clusters based on a minimum citation threshold of 16 per document. These clusters highlight cohesive groups of highly cited papers, indicating thematic similarities and influential works. Here, influential authors with significant citation impact include ”Nayal,” ”Baryannis,” ”Ivanov,” ”Hosseini,” ”Modgil,” ”Zouari,” ”Wu,” and ”Hosseini,” suggesting their substantial contributions.

4.10.4 Organizations

In Figure  12 , 34 significant entities out of 769 met the dual criteria of a minimum document threshold of 5 and a minimum citation threshold of 10, forming 3 clusters. These clusters illustrate collaborative research efforts and the impact of organizations in the academic landscape. It shows institutions like ”Hong Kong Polytechnic University,” ”Neoma Business School,” and ”Indian Institute of Technology Delhi” as notable contributors to these cohesive groups.

4.10.5 Journals

For the bibliographic coupling analysis of sources, 42 significant publication sources out of 196 met the criteria, forming clusters based on a minimum document threshold of 3 per source, as shown in Figure  13 . These clusters reveal patterns in scholarly publishing, highlighting sources that frequently publish related content. Journals such as ”Journal of Computational and Applied Mathematics,” ”IEEE Transactions on Engineering Management,” ”Annals of Operations Research,” and ”International Journal of Production Research” frequently publish related content within their respective clusters.

4.11 Journal Publication Trends

Table  3 presents a comprehensive breakdown of published papers in various source titles from 2015 to 2023. These source titles encompass a range of journals and publications contributing to the field of SCRA. The table clearly shows the publication trends, indicating how many papers were published in each source title each year and their cumulative total. It provides valuable insights into the distribution of research output and the source titles that have been active in this area.

Through these figures and tables, our systematic literature review is empowered to offer an in-depth analysis of the SCRA landscape and provide valuable insights to both the academic and professional communities in this dynamic field.

5 AI Techniques for Supply Chain Risk Assessment

In recent years, the application of Machine Learning (ML) methods in Supply Chain Risk Assessment (SCRA) has garnered substantial attention due to its potential to enhance decision-making processes and mitigate operational risks. This review paper explores various ML techniques applied in SCRA, highlighting their methodologies, key findings, and contributions.

Han and Zhang, ( 2021 ) employed Factor Analysis and a Backpropagation Neural Network (BPNN) to identify high and extremely high-risk incentives in sustainable supply chains, offering valuable insights for risk assessment. Jianying et al., ( 2021 ) optimized a BPNN using Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), enhancing its predictive accuracy, which can be instrumental in the fresh grape industry’s sustainable supply chain management.

Belhadi et al., 2021b utilized a hybrid Ensemble Machine Learning (EML) approach involving the Gama Test, Rotation Forest (RotF), and Logit Boosting (LB) algorithms, achieving excellent performance with the RotF-LB model, contributing to credit risk assessment in the context of Agriculture 4.0. Wu et al., 2022c improved credit risk prediction in agricultural SCF using a GA-BPNN Credit Risk Model, attaining a risk prediction accuracy above 0.92. Zhang et al., ( 2015 ) compared different credit risk assessment models for SCF, highlighting the superiority of the Support Vector Machine (SVM) model with an accuracy of 93.65%. Lei et al., ( 2023 ) demonstrated the effectiveness of the Support Vector Machine and Slime Mould Algorithm (SVM-SMA) model for financial risk assessment, with results showing Precision: 85.38%, F-score: 63%, and TNR: 72%.

Yin et al., ( 2022 ) introduced a Convolutional Neural Network (CNN) model for early warning of SCF risks, achieving an optimal accuracy at 200 iterations and a comprehensive accuracy of 94.7%. Chen et al., ( 2021 ) employed a Long Short-Term Memory (LSTM) network model to forecast oil import risk, achieving superior forecasting accuracy compared to other models, offering a more effective risk assessment solution for oil import scenarios. Janjua et al., ( 2023 ) developed a real-time social media data assessment tool for SCRM with the Bi-Directional Long Short-Term Memory Conditional Random Field (Bi-LSTM CRF) model, resulting in improved precision, recall, and F1 scores compared to baseline models. Yao et al., ( 2022 ) showcased the power of combining multiple ranking information and an ensemble feature selection method for enterprise credit risk assessment, obtaining the best prediction results. Aboutorab et al., ( 2022 ) demonstrated the application of Reinforcement Learning (RL) in proactive risk identification, achieving high accuracy in identifying various risks. Kosasih et al., ( 2022 ) presented a tensorization-based Neuro-Symbolic model that outperformed existing models and unveiled hidden supply chain risks.

Wang et al., 2022a introduced a novel cost-sensitive learning Random Forest model (CSL-RF) equipped with an imbalance sampling strategy to predict SMEs credit risk in SCF, achieving exceptional accuracy (97.%) and precision (1.00). Wang et al., ( 2021 ) addressed the small sample size issue in SME credit risk forecasting by proposing an Adaptive Heterogeneous Multiview Graph Learning (AH-MGL) method, which effectively improved interpretability and demonstrated robustness with an accuracy of 87.44%. Zhu et al., ( 2016 ) introduced the Random Subspace-Real AdaBoost (RS-RAB) method, showcasing its effectiveness in credit risk prediction, and emphasized the performance of ensemble methods against individual machine learning techniques, shedding light on their predictive capabilities. Bassiouni et al., ( 2023 ) presented a unique case study utilizing Deep Learning (DL) methodologies, such as Recurrent Neural Networks (RNN) and Temporal Convolution Neural Networks (TCN), to predict the exportability of shipments during the COVID-19 pandemic, significantly improving classification accuracy.

Zhu et al., ( 2019 ) introduced an enhanced hybrid Ensemble Machine Learning (EML) approach (RS-MultiBoosting) for forecasting SMEs’ credit risk, demonstrating its superiority, especially with small datasets. Athaudage et al., ( 2022 ) proposed a Random Forest (RF) regression model for crude oil price analysis during disease outbreaks, offering improved forecasting capabilities with high accuracy. Liu and Huang, ( 2020 ) combined AdaBoosting and SVM with Adaptive Mutation Particle Swarm Optimization (AM-PSO) and noise reduction techniques for risk assessment in SCF, resulting in higher credit assessment accuracy. Zhu et al., ( 2017 ) analyzed various EML methods, highlighting the effectiveness of the Random Subspace Boosting (RS-Boosting) approach in SME credit risk prediction.

Fu et al., ( 2022 ) explored the application of BPNN, SVM, and GA in risk assessment. The results highlighted the superiority of the BP-GA model in terms of classification accuracy, emphasizing the importance of innovative approaches to risk assessment. Zhang, ( 2022 ) introduced the Fuzzy Support Vector Machine Prediction Model (FSVM) for enterprise SCRA, demonstrating higher accuracy compared to traditional models, including SVM and Decision Trees (DT), and the neural network model Temporal-Spatial Fuzzy Neural Network (T-SFNN). This research addressed the need for company-focused SCRM solutions, especially in the context of macroeconomic environments and epidemics. Pan and Miao, ( 2023 ) presented a DL-based BPNN model for SCRA, showcasing exceptional generalization ability and highlighting its potential as an effective risk assessment method. Yang et al., ( 2021 ) introduced the Least Absolute Shrinkage and Selection Operator Logistic Regression (Lasso-Logistic) model, effectively overcoming multicollinearity and offering the best prediction accuracy and discrimination ability.

Zhang et al., ( 2022 ) introduced the DeepRisk model, emphasizing the importance of utilizing both static and dynamic data in credit risk prediction for SMEs in SCF. This model outperformed baseline methods in various credit risk prediction metrics, emphasizing the value of dynamic financing behavioral data. Wang, ( 2021 ) integrated Blockchain and Fuzzy Neural Network algorithms to study credit risk in SME financing within the SCF context. This approach significantly contributes to credit risk understanding and management within SCF. Li, ( 2022 ) applied BPNN and LR analysis to predict investment risk in SCM, demonstrating high prediction accuracy, which can be instrumental in improving SCM and the sustainable development of enterprises. Bodendorf et al., ( 2023 ) introduced a DL model that exhibited high predictive performance and enhanced interpretability for managers, answering the ”why” question. This approach represented an extension to SCRM research.

Liu et al., ( 2022 ) utilized a hybrid model chain, incorporating eXtreme Gradient Boosting (XGBoost), Synthetic Minority Oversampling Technique for Nominal and Continuous (SMOTENC), and Random Forest (RF), to identify and control credit risk in financial institutions. This approach contributed to developing integrated models that address interpretability and accuracy issues. Rao and Li, ( 2022 ) implemented various ML algorithms, including Logistic Regression (LR), Decision Trees (DT), and integrated Logistic Regression-Random Subspace (LR-RS) and Decision Trees-Random Subspace (DT-RS) methods to enhance risk assessment and behavior prediction in blockchain SCF. The Logistic Regression-Random Subspace (LR-RS) model demonstrated practicality and superiority in credit risk assessment. Zheng et al., ( 2023 ) introduced a Federated Learning (FL) method to address supply chain risk prediction. FL enabled supply chain members with smaller and imbalanced datasets to leverage collective information, enhancing predictive performance. It outperformed other algorithms, demonstrating the potential of collective learning in supply chain risk prediction.

Handfield et al., ( 2020 ) presented a method for predicting factory risks in Low-Cost Countries (LCCs) with a five-year planning horizon, enabling proactive risk assessment before sourcing decisions. This approach addressed the geographic-specific risks associated with supplier factories and informed sourcing allocation decisions. Wong et al., ( 2022 ) highlighted the significance of AI-driven risk management in enhancing Supply Chain Agility and Re-engineering Capabilities (RP). This study explored the potential of artificial intelligence in SME risk assessment and its impact on business continuity. Wang and Song, ( 2022 ) utilized the Interval Type 2 Fuzzy Neural Network (IT2FNN) method, effectively addressing periodic deception and exhibiting a high comprehensive accuracy rate. The research combined fuzzy logic systems and artificial neural networks to assess e-commerce credit risk.

Nezamoddini et al., ( 2020 ) introduced a novel GA integrated with an Artificial Neural Network (ANN) to optimize supply chain decision-making, emphasizing the robustness and profitability improvements the proposed technique offers. Zhang, ( 2022 ) presented an SVM based on a Fuzzy Model (FSVM) for enterprise SCRA. This method’s accuracy surpassed traditional models, contributing to the effectiveness of complex enterprise SCRA. Xia et al., ( 2023 ) explored ML methods for manufacturing SCF credit risk assessment. RF emerged as the best-performing model, and the research focused on analyzing the unique environment of Small and Medium-Sized Manufacturing Enterprises (SMMEs) in China. Zhao and Li, ( 2022 ) examined the use of SVM and BPNN algorithms for measuring risk in SCF, aiming to address the financing dilemma of SMEs.

Li and Fu, ( 2022 ) highlighted the superiority of the Principal Component Analysis-Genetic Algorithm-Support Vector Machine (PCA-GA-SVM) model for credit risk prediction in the specific context of SCF accounts receivable mode. The model exhibited better performance in terms of accuracy and error rates. Duan et al., ( 2021 ) employed a BPNN optimized by the GA model. This approach offered improved fitting effects, higher prediction precision, and faster convergence than conventional models. Zhang et al., ( 2021 ) introduced the Firefly Algorithm Support Vector Machine (FA-SVM) for financial credit risk assessment in supply chains. The FA-SVM enhanced classification efficiency and reduced error rates, making it more practical.

Luo et al., ( 2022 ) used SVM with improved optimization techniques, such as Dynamic Mutation Particle Swarm Optimization (DM-PSO), to optimize parameters. The integration of AdaBoost as a weak classifier resulted in an integrated model with superior performance. Dang et al., ( 2022 ) constructed a BPNN and incorporated blockchain technology for credit risk evaluation in SCF. The research demonstrated the effectiveness of deep learning and blockchain in supply chain financial prediction and management. Hosseini and Barker, ( 2016 ) proposed a Bayesian Network (BN) model to evaluate and select suppliers, considering criteria falling into primary, green, and resilience categories. This model quantified supplier resilience in absorptive, adaptive, and restorative capacities, addressing the importance of resilience in the supplier selection process. Gao et al., ( 2022 ) explored various ML models, including Autoregressive Integrated Moving Average (ARIMA) and Multilayer Perceptron-Long Short-Term Memory (MLP-LSTM), for commodity demand prediction. The research highlighted the effectiveness of the Autoregressive Mixture Density Network (AR-MDN) commodity prediction model and the Improved Particle Swarm Optimization (IPSO) algorithm.

Wei, ( 2022 ) introduced a Machine Learning-based Linear Regression Algorithm (ML-LRA) for credit risk assessment. This method efficiently reduced the risk of supplier credit risk assessment, contributing to the field of supply chain management. Deiva Ganesh and Kalpana, ( 2022 ) employed text mining to analyze live supply chain-related information from social media platforms, emphasizing the importance of real-time risk identification. Zhang et al., ( 2023 ) presented a Time-Decayed Long Short-Term Memory (TD-LSTM) algorithm for monitoring and interpolating missing data in irregular time series. This algorithmic model enhanced the predictability of datasets with missing data. Baryannis et al., 2019a proposed a Generic Data-Driven Risk Prediction Framework (GDDRPF) for SCRM. The framework emphasized the synergy between Artificial Intelligence (AI) and supply chain experts, considering the trade-off between interpretability and prediction performance.

Collectively, these papers contribute to the evolving field of supply chain risk assessment and management, leveraging various machine learning, optimization, and data-driven approaches to enhance prediction, decision-making, and resilience within supply chains.

5.1 Used ML Models and Key Findings

Table 4 summarizes various research methods or models used in risk prediction and analysis, along with the key findings and associated metrics from different research papers. Each row in the table represents a research paper’s first author, the research method or model used, and the specific findings and metrics achieved. These methods and findings are crucial in understanding risk prediction in various domains.

5.2 Clustering of the ML Models

Table  5 presents a clustering of ML models utilized in SCRA, categorizing them into diverse methodological clusters. It offers a comprehensive view of various approaches employed, including ensemble learning, deep learning, time series models, fuzzy logic models, reinforcement learning, dimensionality reduction, genetic algorithms, Bayesian models, and other innovative approaches used in this domain.

6 Managerial Implications

The findings of this study have significant implications for managerial strategies aimed at enhancing SCRM. Firstly, it underscores the pivotal role of ML and AI in revolutionizing SCRA. Managers are encouraged to explore the adoption of advanced ML models, such as Random Forest, XGBoost, and hybrid approaches, as they have demonstrated substantial potential in improving risk assessment precision. These models can significantly bolster the accuracy of risk evaluation and enhance the overall effectiveness of risk management strategies.

Moreover, the research emphasizes the importance of adaptability and flexibility in the post-COVID era, highlighting the need for proactive supply chain strategies. Managers should prioritize the development of contingency plans that are resilient to potential future disruptions, ensuring the continuity of operations and minimizing the impact of unforeseen events.

The study also underscores the effectiveness of ensemble models, such as Random Forest and XGBoost, along with hybrid approaches in risk assessment. These models can potentially enhance risk prediction precision and should be considered by managers seeking to fortify their SCRM. Time series forecasting techniques, including ARIMA and LSTM, emerge as pivotal in anticipating future risks. Organizations are encouraged to invest in predictive models grounded in historical data, providing them with the foresight required to mitigate supply chain disruptions effectively.

The significance of XAI and ML models cannot be overstated. Ensuring that models offer clear and interpretable results is essential, as this aids decision-makers in comprehending the risk assessment process and its outcomes. The study also emphasizes the importance of real-time data mining and incorporating unconventional data sources, such as social media platforms like Twitter, as essential elements of comprehensive risk identification. Managers must stay vigilant, continually updating their understanding of real-time information.

Blockchain technology emerges as a potent tool for enhancing transparency and traceability within the supply chain. Organizations considering blockchain solutions should evaluate their applications for risk mitigation, ultimately enhancing SCR. SMEs warrant tailored strategies for credit risk assessment, accounting for their unique operational challenges. Implementing strategies aligned with the specific needs and constraints of SMEs is vital. In cases where data privacy is a concern, the adoption of federated learning is recommended. This approach safeguards sensitive information while still enabling collaborative risk prediction. Ensuring data security and privacy compliance should be a top priority.

Hybrid risk assessment methodologies, which integrate various ML models and techniques, offer a holistic view of SCRA. Managers should explore these comprehensive approaches to build robust risk management strategies. Benchmarking different ML models is crucial for identifying the most effective ones for specific risk scenarios. The study underscores the significance of continuous performance evaluation and improvement, ensuring that risk assessment strategies evolve with changing circumstances. Integrating ML-based risk assessment into existing supply chain workflows and decision-making processes is pivotal to ensure seamless implementation. Managers should explore ways to align these technologies with their current operational strategies. Recognizing the importance of employee training in data analysis, ML, and AI is fundamental. Managers should invest in skill development to empower their supply chain professionals to harness these technologies effectively. Collaboration and knowledge sharing within organizations and across industries are encouraged to collectively build robust risk management strategies. Such partnerships have the potential to enhance risk resilience significantly.

Lastly, understanding and adhering to data protection and privacy regulations is critical when deploying AI and ML for risk assessment. Managers should ensure that their risk management practices are fully aligned with the legal requirements to avoid any compliance issues. The study offers a comprehensive roadmap for organizations aiming to fortify their SCRM strategies in an ever-evolving landscape.

7 Challenges and Limitations

Integrating AI in SCRA brings forth several challenges and limitations that researchers and practitioners must address. These challenges encompass the following:

Data Availability and Quality: AI models in SCRA rely heavily on data. However, ensuring the availability and quality of data remains a persistent challenge. Supply chain data often comes from various sources, which can be fragmented, incomplete, or outdated. Researchers must explore ways to consolidate, cleanse, and augment data to enhance the effectiveness of AI models.

Interpretability and Explainability: Many AI techniques, such as deep learning and neural networks, are often considered as ”black-box” models. This lack of interpretability and explainability is a significant limitation. In SCRA, stakeholders require insights into why a model makes a particular prediction or decision. Developing AI models that are not only accurate but also interpretable is crucial for their adoption in real-world supply chain management.

Dynamic and Evolving Risk Factors: The supply chain environment is dynamic and continuously evolving. New risk factors emerge, and existing ones change over time. AI models may struggle to adapt to these dynamic conditions. Researchers need to explore techniques for real-time risk assessment and adaptive models that can respond to changing risk profiles.

Integration with Existing Systems: Integrating AI solutions into existing SCM systems can be challenging. Organizations often rely on legacy systems, and ensuring compatibility and seamless integration can pose hurdles. Research in this area should focus on creating modular AI solutions that can be integrated into diverse systems with minimal disruption.

8 Future Research Directions

In light of the challenges and limitations, the future research directions in the field of AI-based supply chain risk assessment are multi-faceted and hold the potential to advance the state-of-the-art:

AI Model Explainability: Emphasize the development of AI models that deliver accurate predictions and offer comprehensive transparency and interpretability. Techniques like LIME (Local Interpretable Model-Agnostic Explanations) and SHAP (SHapley Additive exPlanations) serve as pivotal approaches to enhance the interpretability of AI models. By employing these methods, the goal is to demystify the decision-making process of AI models, enabling stakeholders to comprehend the rationale behind specific predictions or decisions. This transparency enhances trust and facilitates more effective integration of AI-driven insights into supply chain risk management strategies.

Hybrid AI Approaches: Combining multiple AI techniques, such as machine learning and expert systems, can enhance the robustness of supply chain risk assessment. Future research should explore hybrid AI models that leverage the strengths of different approaches to improve prediction accuracy and reliability.

Data Quality and Accessibility: Addressing data quality challenges involves multifaceted strategies. Developing robust methods for data cleansing, augmentation, and ensuring high-quality data remains a critical focus. Beyond that, initiatives should be aimed at democratizing access to standardized supply chain data repositories. Creating and maintaining such repositories enables stakeholders across the supply chain spectrum to access reliable and consistent data. This increased accessibility fosters better-informed decision-making processes and promotes using standardized data sets for improved AI model training and validation.

Real-time Risk Assessment: Investigating AI models capable of real-time risk assessment will be pivotal. These models should adapt to emerging risks and provide timely alerts or recommendations for risk mitigation.

Integration Frameworks: Research on creating integration frameworks for AI-based supply chain risk assessment is essential. These frameworks should facilitate easy integration into existing supply chain management systems, ensuring a smooth transition to AI-driven risk assessment.

Cross-Domain Application: Expanding the application of AI in risk assessment to various domains within the supply chain, including logistics, procurement, and production, offers opportunities for comprehensive risk management.

Adaptation to Pandemic Dynamics (e.g., COVID-19): Investigate AI models capable of adapting to pandemic-induced disruptions within supply chains. Focus on understanding the unique dynamics and challenges posed by events like COVID-19, aiming to develop AI-driven strategies that enable resilience and agility in such crisis scenarios.

Ethical and Legal Considerations: As AI plays an increasingly significant role in decision-making, researchers should also address ethical and legal considerations. This includes issues related to data privacy, bias, and fairness in AI models.

By addressing these challenges and pursuing these research directions, the field of AI-based supply chain risk assessment can make significant strides toward more effective and reliable risk management in an ever-evolving global supply chain landscape.

9 Conclusions

In this extensive review, we delved into the supply chain risk assessment domain and examined the complex challenges that arise when integrating artificial intelligence, especially machine learning. Our thorough analysis encompassed a substantial pool of 1,717 SCRA papers, ultimately narrowing our focus to a refined subset of 48 papers. This endeavor has yielded significant insights and key findings, making substantial contributions to the SCRA literature.

Our primary contributions can be summarized by addressing the four key research questions:

Our review thoroughly analyzes the existing state of research concerning the application of AI and ML techniques in SCRM. We have meticulously assessed a substantial body of literature to extract meaningful insights, specifically 48 selected articles.

Through our analysis, we have identified the AI and ML techniques that are most commonly utilized in the domain of SCRA. This adds clarity to the methodologies employed and underscores the trends shaping the field.

Our review highlights the key findings and trends prevalent in the literature, shedding light on the progress, developments, and major areas of focus within the realm of AI and ML in SCRM.

We have identified critical research gaps and proposed future directions to guide further explorations in this field, ensuring the continued evolution of SCRA methods.

Despite these contributions, our research has some limitations. Our data primarily originated from Google Scholar and Web of Science rather than Scopus, potentially introducing biases. Our focus on papers published between 2014 and 2023 introduces a limitation, as it may omit earlier relevant research that could provide valuable historical context. Additionally, our restriction to English-language papers may have omitted valuable non-English research.

Looking forward, several promising avenues for future research emerge in the domain of SCRA using AI. These directions encompass operationalizing interconnectedness, transformability, and sharing within SCRA frameworks, investigating evolving ICT roles in prediction and response, revisiting the cost implications of resilience enhancement, and exploring emerging technologies like blockchain. These endeavors are expected to enhance risk assessment effectiveness in the post-COVID era. Reviewers should adopt a comprehensive approach to gain deeper insights into this field by expanding their search to diverse databases, including non-English and grey literature sources, using snowballing techniques, collaborating with experts from related disciplines, leveraging ML tools for data analysis, and staying up-to-date with the latest research. Combining various search methods and expert opinions, such a multidisciplinary strategy is essential for uncovering valuable insights and emerging trends in this dynamic and critical field.

Our SLR, driven by these four primary research questions, has contributed valuable insights into the AI/ML application in SCRA, identified common techniques, outlined key findings and trends, and proposed essential research directions. This work serves as a guide for both researchers and practitioners, facilitating advancements in SCRM.

• Aboutorab et al., (2022) Aboutorab, H., Hussain, O. K., Saberi, M., and Hussain, F. K. (2022). A reinforcement learning-based framework for disruption risk identification in supply chains. Future Generation Computer Systems , 126:110–122.
• Agca et al., (2022) Agca, S., Babich, V., Birge, J. R., and Wu, J. (2022). Credit Shock Propagation Along Supply Chains: Evidence from the CDS Market. Management Science , 68(9):6506–6538. Publisher: INFORMS.
• Araz et al., (2020) Araz, O. M., Choi, T.-M., Olson, D. L., and Salman, F. S. (2020). Data Analytics for Operational Risk Management. Decision Sciences , 51(6):1316–1319. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/deci.12443.
• Athaudage et al., (2022) Athaudage, G. N., Perera, H. N., Sugathadasa, P. R. S., De Silva, M. M., and Herath, O. K. (2022). Modelling the impact of disease outbreaks on the international crude oil supply chain using Random Forest regression. International Journal of Energy Sector Management , ahead-of-print(ahead-of-print).
• (5) Baryannis, G., Dani, S., and Antoniou, G. (2019a). Predicting supply chain risks using machine learning: The trade-off between performance and interpretability. Future Generation Computer Systems , 101:993–1004.
• (6) Baryannis, G., Validi, S., Dani, S., and Antoniou, G. (2019b). Supply chain risk management and artificial intelligence: state of the art and future research directions. International Journal of Production Research , 57(7):2179–2202. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/00207543.2018.1530476.
• Bassiouni et al., (2023) Bassiouni, M. M., Chakrabortty, R. K., Hussain, O. K., and Rahman, H. F. (2023). Advanced deep learning approaches to predict supply chain risks under COVID-19 restrictions. Expert Systems with Applications , 211:118604.
• (8) Belhadi, A., Kamble, S., Jabbour, C. J. C., Gunasekaran, A., Ndubisi, N. O., and Venkatesh, M. (2021a). Manufacturing and service supply chain resilience to the COVID-19 outbreak: Lessons learned from the automobile and airline industries. Technological Forecasting and Social Change , 163:120447.
• (9) Belhadi, A., Kamble, S. S., Mani, V., Benkhati, I., and Touriki, F. E. (2021b). An ensemble machine learning approach for forecasting credit risk of agricultural SMEs’ investments in agriculture 4.0 through supply chain finance. Annals of Operations Research .
• Bode et al., (2011) Bode, C., Wagner, S. M., Petersen, K. J., and Ellram, L. M. (2011). Understanding Responses to Supply Chain Disruptions: Insights from Information Processing and Resource Dependence Perspectives. Academy of Management Journal , 54(4):833–856. Publisher: Academy of Management.
• Bodendorf et al., (2023) Bodendorf, F., Sauter, M., and Franke, J. (2023). A mixed methods approach to analyze and predict supply disruptions by combining causal inference and deep learning. International Journal of Production Economics , 256:108708.
• Chen et al., (2021) Chen, S., Song, Y., Ding, Y., Zhang, M., and Nie, R. (2021). Using long short-term memory model to study risk assessment and prediction of China’s oil import from the perspective of resilience theory. Energy , 215:119152.
• Craighead et al., (2020) Craighead, C. W., Ketchen Jr., D. J., and Darby, J. L. (2020). Pandemics and Supply Chain Management Research: Toward a Theoretical Toolbox*. Decision Sciences , 51(4):838–866. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/deci.12468.
• Dang et al., (2022) Dang, C., Wang, F., Yang, Z., Zhang, H., and Qian, Y. (2022). Evaluating and forecasting the risks of small to medium-sized enterprises in the supply chain finance market using blockchain technology and deep learning model. Operations Management Research , 15(3):662–675.
• Deiva Ganesh and Kalpana, (2022) Deiva Ganesh, A. and Kalpana, P. (2022). Supply chain risk identification: a real-time data-mining approach. Industrial Management & Data Systems , 122(5):1333–1354. Publisher: Emerald Publishing Limited.
• Denyer and Tranfield, (2009) Denyer, D. and Tranfield, D. (2009). Producing a systematic review. In The Sage handbook of organizational research methods , pages 671–689. Sage Publications Ltd, Thousand Oaks, CA.
• Duan et al., (2021) Duan, Y., Mu, C., Yang, M., Deng, Z., Chin, T., Zhou, L., and Fang, Q. (2021). Study on early warnings of strategic risk during the process of firms’ sustainable innovation based on an optimized genetic BP neural networks model: Evidence from Chinese manufacturing firms. International Journal of Production Economics , 242:108293.
• Fu et al., (2022) Fu, W., Zhang, H., and Huang, F. (2022). Internet-based supply chain financing-oriented risk assessment using BP neural network and SVM. PLOS ONE , 17(1):e0262222. Publisher: Public Library of Science.
• Gao et al., (2022) Gao, Q., Xu, H., and Li, A. (2022). The analysis of commodity demand predication in supply chain network based on particle swarm optimization algorithm. Journal of Computational and Applied Mathematics , 400:113760.
• Govindan et al., (2020) Govindan, K., Mina, H., and Alavi, B. (2020). A decision support system for demand management in healthcare supply chains considering the epidemic outbreaks: A case study of coronavirus disease 2019 (COVID-19). Transportation Research Part E: Logistics and Transportation Review , 138:101967.
• Guo et al., (2021) Guo, Y., Fu, Y., Hao, F., Zhang, X., Wu, W., Jin, X., Robin Bryant, C., and Senthilnath, J. (2021). Integrated phenology and climate in rice yields prediction using machine learning methods. Ecological Indicators , 120:106935.
• Han and Zhang, (2021) Han, C. and Zhang, Q. (2021). Optimization of supply chain efficiency management based on machine learning and neural network. Neural Computing and Applications , 33(5):1419–1433.
• Handfield et al., (2020) Handfield, R., Sun, H., and Rothenberg, L. (2020). Assessing supply chain risk for apparel production in low cost countries using newsfeed analysis. Supply Chain Management: An International Journal , 25(6):803–821. Publisher: Emerald Publishing Limited.
• Hosseini and Barker, (2016) Hosseini, S. and Barker, K. (2016). A Bayesian network model for resilience-based supplier selection. International Journal of Production Economics , 180:68–87.
• Ivanov, (2020) Ivanov, D. (2020). Predicting the impacts of epidemic outbreaks on global supply chains: A simulation-based analysis on the coronavirus outbreak (COVID-19/SARS-CoV-2) case. Transportation Research Part E: Logistics and Transportation Review , 136:101922.
• Janiesch et al., (2021) Janiesch, C., Zschech, P., and Heinrich, K. (2021). Machine learning and deep learning. Electronic Markets , 31(3):685–695.
• Janjua et al., (2023) Janjua, N. K., Nawaz, F., and Prior, D. D. (2023). A fuzzy supply chain risk assessment approach using real-time disruption event data from Twitter. Enterprise Information Systems , 17(4):1959652. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/17517575.2021.1959652.
• Jianying et al., (2021) Jianying, F., Bianyu, Y., Xin, L., Dong, T., and Weisong, M. (2021). Evaluation on risks of sustainable supply chain based on optimized BP neural networks in fresh grape industry. Computers and Electronics in Agriculture , 183:105988.
• Katsaliaki et al., (2022) Katsaliaki, K., Galetsi, P., and Kumar, S. (2022). Supply chain disruptions and resilience: a major review and future research agenda. Annals of Operations Research , 319(1):965–1002.
• Kosasih et al., (2022) Kosasih, E. E., Margaroli, F., Gelli, S., Aziz, A., Wildgoose, N., and Brintrup, A. (2022). Towards knowledge graph reasoning for supply chain risk management using graph neural networks. International Journal of Production Research , 0(0):1–17. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/00207543.2022.2100841.
• Kwak et al., (2018) Kwak, D.-W., Seo, Y.-J., and Mason, R. (2018). Investigating the relationship between supply chain innovation, risk management capabilities and competitive advantage in global supply chains. International Journal of Operations & Production Management , 38(1):2–21. Publisher: Emerald Publishing Limited.
• Lalmi et al., (2021) Lalmi, A., Fernandes, G., and Souad, S. B. (2021). A conceptual hybrid project management model for construction projects. Procedia Computer Science , 181:921–930.
• Lei et al., (2023) Lei, Y., Qiaoming, H., and Tong, Z. (2023). Research on Supply Chain Financial Risk Prevention Based on Machine Learning. Computational Intelligence and Neuroscience , 2023:e6531154. Publisher: Hindawi.
• Li et al., (2023) Li, D., Zhi, B., Schoenherr, T., and Wang, X. (2023). Developing capabilities for supply chain resilience in a post-COVID world: A machine learning-based thematic analysis. IISE Transactions , 55(12):1256–1276. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/24725854.2023.2176951.
• Li, (2022) Li, L. (2022). Predicting the Investment Risk in Supply Chain Management Using BPNN and Machine Learning. Wireless Communications and Mobile Computing , 2022:e4340286. Publisher: Hindawi.
• Li and Fu, (2022) Li, M. and Fu, Y. (2022). Prediction of Supply Chain Financial Credit Risk Based on PCA-GA-SVM Model. Sustainability , 14(24):16376. Number: 24 Publisher: Multidisciplinary Digital Publishing Institute.
• Liu et al., (2022) Liu, J., Liu, S., Li, J., and Li, J. (2022). Financial credit risk assessment of online supply chain in construction industry with a hybrid model chain. International Journal of Intelligent Systems , 37(11):8790–8813. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/int.22968.
• Liu and Huang, (2020) Liu, Y. and Huang, L. (2020). Supply chain finance credit risk assessment using support vector machine–based ensemble improved with noise elimination. International Journal of Distributed Sensor Networks , 16(1):1550147720903631. Publisher: SAGE Publications.
• Liu et al., (2019) Liu, Z., Gao, R., Zhou, C., and Ma, N. (2019). Two-period pricing and strategy choice for a supply chain with dual uncertain information under different profit risk levels. Computers & Industrial Engineering , 136:173–186.
• Luo et al., (2022) Luo, S., Xing, M., and Zhao, J. (2022). Construction of Artificial Intelligence Application Model for Supply Chain Financial Risk Assessment. Scientific Programming , 2022:e4194576. Publisher: Hindawi.
• Mohanty et al., (2021) Mohanty, D. K., Parida, A. K., and Khuntia, S. S. (2021). Financial market prediction under deep learning framework using auto encoder and kernel extreme learning machine. Applied Soft Computing , 99:106898.
• Nayal et al., (2021) Nayal, K., Raut, R. D., Queiroz, M. M., Yadav, V. S., and Narkhede, B. E. (2021). Are artificial intelligence and machine learning suitable to tackle the COVID-19 impacts? An agriculture supply chain perspective. The International Journal of Logistics Management , 34(2):304–335. Publisher: Emerald Publishing Limited.
• Naz et al., (2022) Naz, F., Kumar, A., Majumdar, A., and Agrawal, R. (2022). Is artificial intelligence an enabler of supply chain resiliency post COVID-19? An exploratory state-of-the-art review for future research. Operations Management Research , 15(1):378–398.
• Nezamoddini et al., (2020) Nezamoddini, N., Gholami, A., and Aqlan, F. (2020). A risk-based optimization framework for integrated supply chains using genetic algorithm and artificial neural networks. International Journal of Production Economics , 225:107569.
• Ni et al., (2020) Ni, D., Xiao, Z., and Lim, M. K. (2020). A systematic review of the research trends of machine learning in supply chain management. International Journal of Machine Learning and Cybernetics , 11(7):1463–1482.
• Nimmy et al., (2022) Nimmy, S. F., Hussain, O. K., Chakrabortty, R. K., Hussain, F. K., and Saberi, M. (2022). Explainability in supply chain operational risk management: A systematic literature review. Knowledge-Based Systems , 235:107587.
• Pan and Miao, (2023) Pan, W. and Miao, L. (2023). Dynamics and risk assessment of a remanufacturing closed-loop supply chain system using the internet of things and neural network approach. The Journal of Supercomputing , 79(4):3878–3901.
• Punia et al., (2020) Punia, S., Singh, S. P., and Madaan, J. K. (2020). From predictive to prescriptive analytics: A data-driven multi-item newsvendor model. Decision Support Systems , 136:113340.
• Rao and Li, (2022) Rao, Q. and Li, W. (2022). Risk Evaluation and Forecast Behavior Analysis of Supply Chain Financing Based on Blockchain. Wireless Communications and Mobile Computing , 2022:e7668474. Publisher: Hindawi.
• Roukny et al., (2018) Roukny, T., Battiston, S., and Stiglitz, J. E. (2018). Interconnectedness as a source of uncertainty in systemic risk. Journal of Financial Stability , 35:93–106.
• Schroeder and Lodemann, (2021) Schroeder, M. and Lodemann, S. (2021). A Systematic Investigation of the Integration of Machine Learning into Supply Chain Risk Management. Logistics , 5(3):62. Number: 3 Publisher: Multidisciplinary Digital Publishing Institute.
• Shahed et al., (2021) Shahed, K. S., Azeem, A., Ali, S. M., and Moktadir, M. A. (2021). A supply chain disruption risk mitigation model to manage COVID-19 pandemic risk. Environmental Science and Pollution Research .
• Sheng et al., (2021) Sheng, J., Amankwah-Amoah, J., Khan, Z., and Wang, X. (2021). COVID-19 Pandemic in the New Era of Big Data Analytics: Methodological Innovations and Future Research Directions. British Journal of Management , 32(4):1164–1183. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/1467-8551.12441.
• Shi et al., (2022) Shi, S., Tse, R., Luo, W., D’Addona, S., and Pau, G. (2022). Machine learning-driven credit risk: a systemic review. Neural Computing and Applications , 34(17):14327–14339.
• Wang et al., (2021) Wang, C., Yu, F., Zhang, Z., and Zhang, J. (2021). Multiview Graph Learning for Small- and Medium-Sized Enterprises’ Credit Risk Assessment in Supply Chain Finance. Complexity , 2021:e6670873. Publisher: Hindawi.
• (56) Wang, L., Jia, F., Chen, L., and Xu, Q. (2022a). Forecasting smes’ credit risk in supply chain finance with a sampling strategy based on machine learning techniques. Annals of Operations Research , pages 1–33.
• (57) Wang, L., Jia, F., Chen, L., and Xu, Q. (2022b). Forecasting SMEs’ credit risk in supply chain finance with a sampling strategy based on machine learning techniques. Annals of Operations Research .
• Wang and Song, (2022) Wang, L. and Song, H. (2022). E-Commerce Credit Risk Assessment Based on Fuzzy Neural Network. Computational Intelligence and Neuroscience , 2022:e3088915. Publisher: Hindawi.
• Wang, (2021) Wang, Y. (2021). Research on Supply Chain Financial Risk Assessment Based on Blockchain and Fuzzy Neural Networks. Wireless Communications and Mobile Computing , 2021:e5565980. Publisher: Hindawi.
• Wei, (2022) Wei, Y. (2022). A Machine Learning Algorithm for Supplier Credit Risk Assessment Based on Supply Chain Management. International Transactions on Electrical Energy Systems , 2022:e4766597. Publisher: Hindawi.
• Wong et al., (2020) Wong, C. W. Y., Lirn, T.-C., Yang, C.-C., and Shang, K.-C. (2020). Supply chain and external conditions under which supply chain resilience pays: An organizational information processing theorization. International Journal of Production Economics , 226:107610.
• Wong et al., (2022) Wong, L.-W., Tan, G. W.-H., Ooi, K.-B., Lin, B., and Dwivedi, Y. K. (2022). Artificial intelligence-driven risk management for enhancing supply chain agility: A deep-learning-based dual-stage PLS-SEM-ANN analysis. International Journal of Production Research , 0(0):1–21. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/00207543.2022.2063089.
• (63) Wu, J., Zhang, Z., and Zhou, S. X. (2022a). Credit rating prediction through supply chains: A machine learning approach. Production and Operations Management , 31(4):1613–1629.
• (64) Wu, J., Zhang, Z., and Zhou, S. X. (2022b). Credit Rating Prediction Through Supply Chains: A Machine Learning Approach. Production and Operations Management , 31(4):1613–1629. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/poms.13634.
• (65) Wu, Y., Li, X., Liu, Q., and Tong, G. (2022c). The Analysis of Credit Risks in Agricultural Supply Chain Finance Assessment Model Based on Genetic Algorithm and Backpropagation Neural Network. Computational Economics , 60(4):1269–1292.
• Xia et al., (2023) Xia, Y., Xu, T., Wei, M.-X., Wei, Z.-K., and Tang, L.-J. (2023). Predicting Chain’s Manufacturing SME Credit Risk in Supply Chain Finance Based on Machine Learning Methods. Sustainability , 15(2):1087. Number: 2 Publisher: Multidisciplinary Digital Publishing Institute.
• Xu and Jackson, (2019) Xu, C. and Jackson, S. A. (2019). Machine learning and complex biological data. Genome Biology , 20(1):76.
• Xu et al., (2020) Xu, S., Zhang, X., Feng, L., and Yang, W. (2020). Disruption risks in supply chain management: a literature review based on bibliometric analysis. International Journal of Production Research , 58(11):3508–3526. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/00207543.2020.1717011.
• Yang et al., (2023) Yang, M., Lim, M. K., Qu, Y., Ni, D., and Xiao, Z. (2023). Supply chain risk management with machine learning technology: A literature review and future research directions. Computers & Industrial Engineering , 175:108859.
• Yang et al., (2021) Yang, Y., Chu, X., Pang, R., Liu, F., and Yang, P. (2021). Identifying and Predicting the Credit Risk of Small and Medium-Sized Enterprises in Sustainable Supply Chain Finance: Evidence from China. Sustainability , 13(10):5714. Number: 10 Publisher: Multidisciplinary Digital Publishing Institute.
• Yao et al., (2022) Yao, G., Hu, X., and Wang, G. (2022). A novel ensemble feature selection method by integrating multiple ranking information combined with an SVM ensemble model for enterprise credit risk prediction in the supply chain. Expert Systems with Applications , 200:117002.
• Yin et al., (2022) Yin, L.-L., Qin, Y.-W., Hou, Y., and Ren, Z.-J. (2022). A Convolutional Neural Network-Based Model for Supply Chain Financial Risk Early Warning. Computational Intelligence and Neuroscience , 2022:e7825597. Publisher: Hindawi.
• Zhang et al., (2020) Zhang, F., Wu, X., Tang, C. S., Feng, T., and Dai, Y. (2020). Evolution of Operations Management Research: from Managing Flows to Building Capabilities. Production and Operations Management , 29(10):2219–2229. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/poms.13231.
• Zhang et al., (2021) Zhang, H., Shi, Y., Yang, X., and Zhou, R. (2021). A firefly algorithm modified support vector machine for the credit risk assessment of supply chain finance. Research in International Business and Finance , 58:101482.
• Zhang et al., (2015) Zhang, L., Hu, H., and Zhang, D. (2015). A credit risk assessment model based on SVM for small and medium enterprises in supply chain finance. Financial Innovation , 1(1):14.
• Zhang et al., (2023) Zhang, W., Lim, M. K., Yang, M., Li, X., and Ni, D. (2023). Using deep learning to interpolate the missing data in time-series for credit risks along supply chain. Industrial Management & Data Systems , 123(5):1401–1417. Publisher: Emerald Publishing Limited.
• Zhang et al., (2022) Zhang, W., Yan, S., Li, J., Tian, X., and Yoshida, T. (2022). Credit risk prediction of SMEs in supply chain finance by fusing demographic and behavioral data. Transportation Research Part E: Logistics and Transportation Review , 158:102611.
• Zhang, (2022) Zhang, X. (2022). Enterprise Supply Chain Risk Assessment Based on the Support Vector Machine Algorithm and Fuzzy Model. Security and Communication Networks , 2022:e3692628. Publisher: Hindawi.
• Zhao and Li, (2022) Zhao, J. and Li, B. (2022). Credit risk assessment of small and medium-sized enterprises in supply chain finance based on SVM and BP neural network. Neural Computing and Applications , 34(15):12467–12478.
• Zheng et al., (2023) Zheng, G., Kong, L., and Brintrup, A. (2023). Federated machine learning for privacy preserving, collective supply chain risk prediction. International Journal of Production Research , 0(0):1–18. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/00207543.2022.2164628.
• Zhu et al., (2016) Zhu, Y., Xie, C., Wang, G.-J., and Yan, X.-G. (2016). Predicting China’s SME Credit Risk in Supply Chain Finance Based on Machine Learning Methods. Entropy , 18(5):195. Number: 5 Publisher: Multidisciplinary Digital Publishing Institute.
• Zhu et al., (2017) Zhu, Y., Xie, C., Wang, G.-J., and Yan, X.-G. (2017). Comparison of individual, ensemble and integrated ensemble machine learning methods to predict China’s SME credit risk in supply chain finance. Neural Computing and Applications , 28(1):41–50.
• Zhu et al., (2019) Zhu, Y., Zhou, L., Xie, C., Wang, G.-J., and Nguyen, T. V. (2019). Forecasting SMEs’ credit risk in supply chain finance with an enhanced hybrid ensemble machine learning approach. International Journal of Production Economics , 211:22–33.

Appendix A Co-authorship Analysis Clusters

Appendix b co-citation analysis clusters, b.1 cited references, b.2 authors, appendix c bibliographic coupling clusters, c.1 authors, c.2 countries, c.3 documents, c.4 organizations.

A systematic review of the research trends of machine learning in supply chain management

• Original Article
• Published: 21 December 2019
• Volume 11 , pages 1463–1482, ( 2020 )

Cite this article

• Zhi Xiao 1 &
• Ming K. Lim   ORCID: orcid.org/0000-0003-0809-9431 2 , 3  

8069 Accesses

89 Citations

1 Altmetric

Explore all metrics

Research interests in machine learning (ML) and supply chain management (SCM) have yielded an enormous amount of publications during the last two decades. However, in the literature, there was no systematic examination on the research development in the discipline of ML application, in particular in SCM. Therefore, this study was carried out to present the latest research trends in the discipline by analyzing the publications between 1998/01/01 and 2018/12/31 in five major databases. The quantitative analysis of 123 shortlisted articles showed that ML applications in SCM were still in a developmental stage since there were not enough high-yielding authors to form a strong group force in the research of ML applications in SCM and their publications were still at a low level; even though 10 ML algorithms were found to be frequently used in SCM, the use of these algorithms were unevenly distributed across the SCM activities most frequently reported in the articles of the literature. The aim of this study is to provide a comprehensive view of ML applications in SCM, working as a reference for future research directions for SCM researchers and application insight for SCM practitioners.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Supply chain disruptions and resilience: a major review and future research agenda

Artificial intelligence-driven innovation for enhancing supply chain resilience and performance under the effect of supply chain dynamism: an empirical investigation

Predictive big data analytics for supply chain demand forecasting: methods, applications, and research opportunities

Aburto L, Weber R (2007) Improved supply chain management based on hybrid demand forecasts. Appl Soft Comput 7(1):136–144

Article   Google Scholar  

Aksoy A, Öztürk N (2011) Supplier selection and performance evaluation in just-in-time production environments. Expert Syst Appl 38(5):6351–6359

Alfian G, Rhee J, Ahn H, Lee J, Farooq U, Ijaz M, Syaekhoni M (2017) Integration of RFID, wireless sensor networks, and data mining in an e-pedigree food traceability system. J Food Eng 212:65–75

Arunraj N, Ahrens D (2015) A hybrid seasonal autoregressive integrated moving average and quantile regression for daily food sales forecasting. Int J Prod Econ 170:321–335

Becker T, Illigen C, McKelvey B, Hülsmann M, Windt K (2016) Using an agent-based neural-network computational model to improve product routing in a logistics facility. Int J Prod Econ 174:156–167

Bhattacharya A, Kumar S, Tiwari M, Talluri S (2014) An intermodal freight transport system for optimal supply chain logistics. Trans Res Part C 38:73–84

Bousqaoui H, Achchab S (2017) Tikito K Machine learning applications in supply chains: an emphasis on neural network applications. In: 2017 3rd international conference of cloud computing technologies and applications (CloudTech). IEEE, pp 1–7

Bowling M, Fürnkranz J, Graepel T, Musick R (2006) Machine learning and games. Mach Learn 63(3):211–215

Brandenburg M, Govindan K, Sarkis J, Seuring S (2014) Quantitative models for sustainable supply chain management: developments and directions. Eur J Oper Res 233(2):299–312

Article   MathSciNet   MATH   Google Scholar  

Breaugh J (2008) Important considerations in using statistical procedures to control for nuisance variables in non-experimental studies. Hum Resource Manage Rev 18(4):282–293

Carbonneau R, Laframboise K, Vahidov R (2008) Application of machine learning techniques for supply chain demand forecasting. Eur J Oper Res 184(3):1140–1154

Article   MATH   Google Scholar  

Carbonneau R, Vahidov R, Laframboise K (2007) Machine learning-based demand forecasting in supply chains. Int J Intell Inf Technol (IJIIT) 3(4):40–57

Chae B (2015) Insights from hashtag supplychain and twitter analytics: considering twitter and twitter data for supply chain practice and research. Int J Prod Econ 16(5):247–259

Chatzidimitriou K, Symeonidis A, Kontogounis I, Mitkas P (2008) Agent Mertacor: a robust design for dealing with uncertainty and variation in SCM environments. Expert Syst Appl 35(3):591–603

Chen M, Tai T, Hung T (2012) Component selection system for green supply chain. Expert Syst Appl 39(5):5687–5701

Cheng J, Chen H, Lin Y (2010) A hybrid forecast marketing timing model based on probabilistic neural network, rough set and C.45. Expert Syst Appl 37(3):1814–1820

Chi H, Ersoy O, Moskowitz H, Ward J (2007) Modeling and optimizing a vendor managed replenishment system using machine learning and genetic algorithms. Eur J Oper Res 180(1):174–193

Chiu M, Lin G (2004) Collaborative supply chain planning using the artificial neural network approach. J Manuf Technol Manage 15(8):787–796

Choy K, Lee W, Lo V (2003) Design of an intelligent supplier relationship management system: a hybrid case based neural network approach. Expert Syst Appl 24(2):225–237

Choy K, Lee W, Lo V (2002) An intelligent supplier management tool for benchmarking suppliers in outsource manufacturing. Expert Syst Appl 22(3):213–224

Christoph K. (2015) The most important algorithms. https://www3.risc.jku.at/people/ckoutsch/stuff/e_algorithms.html

Chung W, Ho C, Wong K, Soon P (2007) An ANN-based DSS system for quality assurance in production network. J Manuf Technol Manage 18(7):836–857

Ćirović G, Pamučar D, Božanić D (2014) Green logistic vehicle routing problem: routing light delivery vehicles in urban areas using a neuro-fuzzy model. Expert Syst Appl 41(9):4245–4258

Cruz J, Wishart D (2007) Applications of machine learning in cancer prediction and prognosis. Cancer Inf 2:59–77

Google Scholar  

De’Ath G (2007) Boosted trees for ecological modeling and prediction. Ecology 88(1):243–251

Ciccio C, Aa H, Cabanillas C, Mendling J, Prescher J (2016) Detecting flight trajectory anomalies and predicting diversions in freight transportation. Decis Support Syst 88:1–17

Dubey R, Gunasekaran A, Papadopoulos T, Childe S, Shibin K, Wamba S (2017) Sustainable supply chain management: framework and further research directions. J Clean Prod 142:1119–1130

Efendigil T, Önüt S (2012) An integration methodology based on fuzzy inference systems and neural approaches for multi-stage supply-chains. Comput Ind Eng 62(2):554–569

Efendigil T, Önüt S, Kongar E (2008) A holistic approach for selecting a third-party reverse logistics provider in the presence of vagueness. Comput Ind Eng 54(2):269–287

Estelles-Lopez L et al (2017) An automated ranking platform for machine learning regression models for meat spoilage prediction using multi-spectral imaging and metabolic profiling. Food Res Int 99:206–215

Fallahpour A, Wong K, Olugu E, Musa S (2017) A predictive integrated genetic-based model for supplier evaluation and selection. Int J Fuzzy Syst 19(4):1041–1057

Article   MathSciNet   Google Scholar  

Fan X, Zhang S, Wang L, Yang Y, Hapeshi K (2013) An evaluation model of supply chain performances using 5DBSC and LMBP neural network algorithm. J Bionic Eng 10(3):383–395

Feng X, Xiao Z, Zhong B, Qiu J, Dong Y (2018) Dynamic ensemble classification for credit scoring using soft probability. Appl Soft Comput 65:139–151

Freeman D (2000) Alternative panel estimates of alcohol demand, taxation, and the business cycle. Southern Econ J 67(2):325–344

Gao L, Shen G, Wang S (2010) Intelligent scheduling model and algorithm for manufacturing. Prod Plan Control 11(3):234–243

García S, Fernández A, Luengo J, Herrera F (2009) A study of statistical techniques and performance measures for genetics-based machine learning: accuracy and interpretability. Soft Comput 13(10):959

Ghasri M, Maghrebi M, Rashidi T, Waller S (2016) Hazard-based model for concrete pouring duration using construction site and supply chain parameters. Autom Constr 71:283–293

Giannakis M, Papadopoulos T (2016) Supply chain sustainability: a risk management approach. Int J Prod Econ 171:455–470

Golmohammadi D, Creese R, Valian H, Kolassa J (2009) Supplier selection based on a neural network model using genetic algorithm. IEEE Trans Neural Netw 20(9):1504–1519

Gumus A, Guneri A, Ulengin F (2010) A new methodology for multi-echelon inventory management in stochastic and neuro-fuzzy environments. Int J Prod Econ 128(1):248–260

Guo X, Yuan Z, Tian B (2009) Supplier selection based on hierarchical potential support vector machine. Expert Syst Appl 36(3):6978–6985

Gupta S, Keen M, Shah A, Verdier G (2017) International monetary fund (2017) Digital revolutions in public finance. International Monetary Fund, Washington

Ha S, Krishnan R (2008) A hybrid approach to supplier selection for the maintenance of a competitive supply chain. Expert Syst Appl 34(2):1303–1311

He X, Ai X, Jing Y, Liu Y (2016) Partner selection of agricultural products supply chain based on data mining. Concurrency Comput 28(4):1246–1256

Hilbert M (2016) Big data for development: a review of promises and challenges. Dev Policy Rev 34(1):135–174

Hong G, Ha S (2008) Evaluating supply partner’s capability for seasonal products using machine learning techniques. Comput Ind Eng 54(4):721–736

Hosseini S, Khaled A (2016) A hybrid ensemble and AHP approach for resilient supplier selection. J Intell Manuf 30(1):207–228

Huang G, Zhu Q, Siew C (2006) Extreme learning machine: theory and applications. Neurocomputing 70(1–3):489–501

Jaipuria S, Mahapatra S (2014) An improved demand forecasting method to reduce bullwhip effect in supply chains. Expert Syst Appl 41(5):2395–2408

Kaelbling L, Littman M, Moore A (1996) Reinforcement learning: a survey. J Artif Intell Res 4:237–285

Keller T, Thiesse F, Fleisch E (2014) Classification models for RFID-based real-time detection of process events in the supply chain. ACM Trans Manage Inf Syst 5(4):1–30

Kiekintveld C, Jain M, Tsai J, Pita J, Ordóñez F, Tambe M (2009) Computing optimal randomized resource allocations for massive security games. In: Proceedings of the 8th international conference on autonomous agents and multiagent systems. International Foundation for Autonomous Agents and Multiagent Systems, pp 689–696

Kim M (2012) Ensemble learning with support vector machines for bond rating. J Intell Inf Syst 18(2):29–45

Ko T, Lee J, Cho H, Cho S, Lee W, Lee M (2017) Machine learning-based anomaly detection via integration of manufacturing, inspection and after-sales service data. Ind Manage Data Syst 117(5):927–945

Kuo R, Chen J (2004) A decision support system for order selection in electronic commerce based on fuzzy neural network supported by real-coded genetic algorithm. Expert Syst Appl 26(2):141–154

Kuo RJ, Wang YC, Tien FC (2010) Integration of artificial neural network and MADA methods for green supplier selection. J Clean Prod 18(12):1161–1170

Kuo R, Xue K (1998) A decision support system for sales forecasting through fuzzy neural networks with asymmetric fuzzy weights. Decis Support Syst 24(1):105–126

Kwon I, Kim C, Jun J, Lee J (2008) Case-based myopic reinforcement learning for satisfying target service level in supply chain. Expert Syst Appl 35(1–2):389–397

Lau H, Hui I, Chan F, Wong C (2002) Monitoring the supply of products in a supply chain environment: a fuzzy neural approach. Expert Syst 19(4):235–243

Lau H, Lee W, Lau P (2001) Development of an intelligent decision support system for benchmarking assessment of business partners. Benchmarking 8(5):376–395

Lau H, Tsui E, Ning A, Pun K, Chin K, Ip W (2005) A knowledge-based system to support procurement decision. J Knowl Manage 9(1):87–100

Lau R, Zhang W, Xu W (2018) Parallel aspect-oriented sentiment analysis for sales forecasting with big data. Prod Oper Manage 27(10):1775–1794

Le C, Bengio Y, Hinton G (2015) Deep learning. Nature 521(7553):436

Lee C, Ho W, Ho G, Lau H (2011) Design and development of logistics workflow systems for demand management with RFID. Expert Syst Appl 38(5):5428–5437

Lee J, Park S (2005) Intelligent profitable customers segmentation system based on business intelligence tools. Expert Syst Appl 29(1):145–152

Lessmann S, Baesens B, Seow H, Thomas L (2015) Benchmarking state-of-the-art classification algorithms for credit scoring: an update of research. Eur J Oper Res 247(1):124–136

Li H, Sun J, Wu J, Wu X (2012) Supply chain trust diagnosis (SCTD) using inductive case-based reasoning ensemble (ICBRE): the case of general competence trust diagnosis. Appl Soft Comput 12(8):2312–2321

Libbrecht M, Noble W (2015) Machine learning applications in genetics and genomics. Nat Rev Genet 16(6):321

Liu C, Shu T, Chen S, Wang S, Lai K, Gan L (2016) An improved grey neural network model for predicting transportation disruptions. Expert Syst Appl 45:331–340

Lloyd S (1982) Least squares quantization in PCM. IEEE Trans Inf Theory 28(2):129–137

Lo Y, Rensi S, Torng W, Altman R (2018) Machine learning in chemoinformatics and drug discovery. Drug Discov Today 23(8):1538–1546

Louzada F, Ara A (2012) Bagging k-dependence probabilistic networks: an alternative powerful fraud detection tool. Expert Syst Appl 39(14):11583–11592

Lu C, Kao L (2016) A clustering-based sales forecasting scheme by using extreme learning machine and ensembling linkage methods with applications to computer server. Eng Appl Artif Intell 55:231–238

Ma H, Wang Y, Wang K (2018) Automatic detection of false positive RFID readings using machine learning algorithms. Expert Syst Appl 91:442–451

MacQueen J (1967) Some methods for classification and analysis of multivariate observations. Proc fifth Berkeley Symp Math Stat Probab 14:281–297

MathSciNet   MATH   Google Scholar  

Maghrebi M, Monty S, Profes A, Sammut C, Waller S (2015) Feasibility study of automatically performing the concrete delivery dispatching through machine learning techniques. Eng Constr Archit Manage 22(5):573–590

Maleki M, Cruz-Machado V (2013) Supply chain performance monitoring using Bayesian network. Int J Bus Perform Supply Chain Model 5(2):177–196

Manyika J, Chui M, Brown B, Bughin J, Dobbs R, Roxburgh C, Byers A (2011) Big data: The next frontier for innovation, competition, and productivity. http://www.mckinsey.com/mgi/publication/big.data/M GI_big_data_exec_summary.pdf(May, 2011)

Mao D, Wang F, Hao Z, Li H (2018) Credit evaluation system based on blockchain for multiple stakeholders in the food supply chain. Int J Environ Res Public Health 15(8):1027

Marr B (2016) Why everyone must get ready for the 4th industrial revolution https://www.forbes.com/sites/bernardmarr/2016/04/05/why-everyone-must-get-readyfor-4th-dustrial-revolution/#26be9e2f3f90

Martínez A, Schmuck C, Pereverzyev S Jr, Pirker C, Haltmeier M (2018) A machine learning framework for customer purchase prediction in the non-contractual setting. Eur J Oper Res 28(23):1–13

Mercier S, Uysal I (2018) Neural network models for predicting perishable food temperatures along the supply chain. Biosys Eng 171:91–100

Min H (2009) Artificial intelligence in supply chain management: theory and applications. Int J Logistics Res Appl 13(1):13–39

Mirkouei A, Haapala K, Sessions J, Murthy G (2017) A mixed biomass-based energy supply chain for enhancing economic and environmental sustainability benefits: a multi-criteria decision making framework. Appl Energy 206:1088–1101

Mishra N, Singh A (2018) Use of twitter data for waste minimisation in beef supply chain. Ann Oper Res 270(1–2):337–359

Mitchell T (1997) Machine Learning. McGraw-Hill, New York

MATH   Google Scholar  

Mori J, Kajikawa Y, Kashima H, Sakata I (2012) Machine learning approach for finding business partners and building reciprocal relationships. Expert Syst Appl 39(12):10402–10407

Mortazavi A, Arshadi Khamseh A, Azimi P (2015) Designing of an intelligent self-adaptive model for supply chain ordering management system. Eng Appl Artif Intell 37:207–220

Ngai E, Peng S, Alexander P, Moon K (2014) Decision support and intelligent systems in the textile and apparel supply chain: an academic review of research articles. Expert Syst Appl 41(1):81–91

Ni D, Xiao Z, Zhong B, Feng X (2018) Multiple human-behaviour indicators for predicting lung cancer mortality with support vector machine. Sci Rep 8(1):16596

Ning A, Lau H, Zhao Y, Wong T (2009) Fulfillment of retailer demand by using the MDL-optimal neural network prediction and decision policy. IEEE Trans Industr Inf 5(4):495–506

Noroozi A, Mokhtari H, Kamal A (2013) Research on computational intelligence algorithms with adaptive learning approach for scheduling problems with batch processing machines. Neurocomputing 101:190–203

Pan Y, Pavur R, Pohlen T (2016) Revisiting the effects of forecasting method selection and information sharing under volatile demand in SCM applications. IEEE Trans Eng Manage 63(4):377–389

Piendl R, Matteis T, Liedtke G (2019) A machine learning approach for the operationalization of latent classes in a discrete shipment size choice model. Trans Res Part E 121:149–161

Piramuthu S (2008) Adaptive framework for collisions in RFID tag identification. J Inf Knowl Manage 7(1):9–14

Ratner B (2000) A comparison of two popular machine learning methods: Common Pitfalls. DM STAT-1: Online Newsletter about Quantitative Methods in Direct Marketing, 4

Raut R, Priyadarshinee P, Gardas BB, Narkhede BE, Nehete R (2018) The incident effects of supply chain and cloud computing integration on the business performance. Benchmarking 25(8):2688–2722

Rodger J (2014) Application of a fuzzy feasibility bayesian probabilistic estimation of supply chain backorder aging, unfilled backorders, and customer wait time using stochastic simulation with Markov blankets. Expert Syst Appl 41(16):7005–7022

Rohde J (2004) Hierarchical supply chain planning using artificial neural networks to anticipate base-level outcomes. OR Spectrum 26(4):471–492

Shervais S, Shannon TT, Lendaris GG (2003) Intelligent supply chain management using adaptive critic learning. IEEE Trans Syst Man Cybern Part A 33(2):235–244

Singh A, Shukla N, Mishra N (2018) Social media data analytics to improve supply chain management in food industries. Trans Res Part E 114:398–415

Stockheim T, Schwind M, Koenig W (2003) A reinforcement learning approach for supply chain management. Electron News August 5(4):1–3

Sun Z, Choi T, Au K, Yu Y (2008) Sales forecasting using extreme learning machine with applications in fashion retailing. Decis Support Syst 46(1):411–419

Swain A, Cao R (2017) Using sentiment analysis to improve supply chain intelligence. Inf Syst Front 21(2):469–484

Syam N, Sharma A (2018) Waiting for a sales renaissance in the fourth industrial revolution: machine learning and artificial intelligence in sales research and practice. Ind Mark Manage 69:135–146

Tavana M, Fallahpour A, Caprio D, Santos-Arteaga F (2016) A hybrid intelligent fuzzy predictive model with simulation for supplier evaluation and selection. Expert Syst Appl 61:129–144

Thomassey S (2010) Sales forecasts in clothing industry: the key success factor of the supply chain management. Int J Prod Econ 128(2):470–483

Timme S, Williams-Timme C (2003) The real cost of holding inventory. Supply Chain Manage Rev 7(4):30–37

Trapero J, Kourentzes N, Fildes R (2012) Impact of information exchange on supplier forecasting performance. Omega 40(6):738–747

Tseng T, Huang C, Jiang F, Ho J (2006) Applying a hybrid data-mining approach to prediction problems: a case of preferred suppliers prediction. Int J Prod Res 44(14):2935–2954

Vahdani B, Iranmanesh S, Mousavi S, Abdollahzade M (2012) A locally linear neuro-fuzzy model for supplier selection in cosmetics industry. Appl Math Model 36(10):4714–4727

Vaat T, Donk D (2004) Buyer focus: evaluation of a new concept for supply chain integration. Int J Prod Econ 92(1):21–30

Vahdani B, Razavi F, Mousavi S (2015) A high performing meta-heuristic for training support vector regression in performance forecasting of supply chain. Neural Comput Appl 27(8):2441–2451

Vapnik V (2013) The nature of statistical learning theory. Springer Science & business media, Berlin

Wan X, Pekny J, Reklaitis G (2005) Simulation-based optimization with surrogate models—application to supply chain management. Comput Chem Eng 29(6):1317–1328

Werbos P (1975) Experimental implications of the reinterpretation of quantum mechanics. II Nuovo Cimento B 29(1):169–177

Wong W, Guo Z (2010) A hybrid intelligent model for medium-term sales forecasting in fashion retail supply chains using extreme learning machine and harmony search algorithm. Int J Prod Econ 128(2):614–624

Wu D (2009) Supplier selection: a hybrid model using DEA, decision tree and neural network. Expert Syst Appl 36(5):9105–9112

Wu H, Evans G, Bae K (2015) Production control in a complex production system using approximate dynamic programming. Int J Prod Res 54(8):2419–2432

Xia M, Zhang Y, Weng L, Ye X (2012) Fashion retailing forecasting based on extreme learning machine with adaptive metrics of inputs. Knowledge-Based Systems 36253-259

Xiao R, Li J, Chen T (2019) Modeling and intelligent optimization of social collective behavior with online public opinion synchronization. Int J Mach Learn Cybern 10(8):1979–1996

Xie G, Zhao Y, Jiang M, Zhang N (2013) A novel ensemble learning approach for corporate financial distress forecasting in fashion and textiles supply chains. Math Prob Eng 1:1–9

Zhao Y, Chen Q (2014) Online order priority evaluation based on hybrid harmony search algorithm of optimized support vector machines. J Netw 9(4):972–978

Zhang L, Hu H, Zhang D (2015) A credit risk assessment model based on SVM for small and medium enterprises in supply chain finance. Financial Innovation 1(1):14–25

Zhu Y, Xie C, Wang G, Yan X (2016) Comparison of individual, ensemble and integrated ensemble machine learning methods to predict China’s SME credit risk in supply chain finance. Neural Comput Appl 28(S1):41–50

Zuo Y, Kajikawa Y, Mori J (2016) Extraction of business relationships in supply networks using statistical learning theory. Heliyon 2(6):e00123

Download references

Author information

Authors and affiliations.

School of Economics and Business Administration, Chongqing University, Chongqing, 400044, China

Du Ni & Zhi Xiao

College of Mechanical Engineering, Chongqing University, Chongqing, 400044, China

Ming K. Lim

Centre for Business in Society, Coventry University, Coventry, CV1 5FB, UK

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ming K. Lim .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Ni, D., Xiao, Z. & Lim, M.K. A systematic review of the research trends of machine learning in supply chain management. Int. J. Mach. Learn. & Cyber. 11 , 1463–1482 (2020). https://doi.org/10.1007/s13042-019-01050-0

Download citation

Received : 18 June 2019

Accepted : 08 December 2019

Published : 21 December 2019

Issue Date : July 2020

DOI : https://doi.org/10.1007/s13042-019-01050-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Machine learning
• Supply chain management
• Research trends
• Application
• Find a journal
• Publish with us
• Track your research

Machine Learning in Supply Chain Management: A Systematic Literature Review

Document Type : GOL20

• Mohamed-Iliasse Mahraz 1
• Loubna Benabbou 2
• Abdelaziz Berrado 1

1 Research team AMIPS, The Mohammadia School of Engineers, Mohammed V University in Rabat, Morocco

2 Department of management sciences, University of Quebec at Rimouski, Levis, QC, Canada

• Supply chain
• Supply network
• Expert system
• Machine Learning
• Digital Transformation
• Supply Chain Analytics

Volume 9, Issue 4 November 2022 Pages 398-416

How to cite.

• Article View: 2,214
• PDF Download: 3,369

Mahraz, M., Benabbou, L., & Berrado, A. (2022). Machine Learning in Supply Chain Management: A Systematic Literature Review. International Journal of Supply and Operations Management , 9(4), 398-416. doi: 10.22034/ijsom.2021.109189.2279

Mohamed-Iliasse Mahraz; Loubna Benabbou; Abdelaziz Berrado. "Machine Learning in Supply Chain Management: A Systematic Literature Review". International Journal of Supply and Operations Management , 9, 4, 2022, 398-416. doi: 10.22034/ijsom.2021.109189.2279

Mahraz, M., Benabbou, L., Berrado, A. (2022). 'Machine Learning in Supply Chain Management: A Systematic Literature Review', International Journal of Supply and Operations Management , 9(4), pp. 398-416. doi: 10.22034/ijsom.2021.109189.2279

Mahraz, M., Benabbou, L., Berrado, A. Machine Learning in Supply Chain Management: A Systematic Literature Review. International Journal of Supply and Operations Management , 2022; 9(4): 398-416. doi: 10.22034/ijsom.2021.109189.2279