Predicting supply chain risks using machine learning: The trade-off between performance and interpretability

Abstract

Managing supply chain risks has received increased attention in recent years, aiming to shield supply chains from disruptions by predicting their occurrence and mitigating their adverse effects. At the same time, the resurgence of Artificial Intelligence (AI) has led to the investigation of machine learning techniques and their applicability in supply chain risk management. However, most works focus on prediction performance and neglect the importance of interpretability so that results can be understood by supply chain practitioners, helping them make decisions that can mitigate or prevent risks from occurring. In this work, we first propose a supply chain risk prediction framework using data-driven AI techniques and relying on the synergy between AI and supply chain experts. We then explore the trade-off between prediction performance and interpretability by implementing and applying the framework on the case of predicting delivery delays in a real-world multi-tier manufacturing supply chain. Experiment results show that prioritising interpretability over performance may require a level of compromise, especially with regard to average precision scores.

Sustainable supply chain management towards disruption and organizational ambidexterity: A data driven analysis

“…(2019) ; Thomas & Mahanty (2019) ; Mikhail et al (2019) ; Singh et al (2019) ; Scholten et al (2019) ; Bevilacqua et al (2019) ; Rajesh (2016) ; Hosseini & Ivanov (2020) ; Behzadi et al (2020) ; Lohmer et al (2020) ; Li & Zobel (2020) . I19 Supply chain risk management SC risk management refers to the coordinated approach among the members of a supply chain for identifying and managing supply chain risk in order to reduce supply chain vulnerability Jüttner et al (2003); Munir et al (2020) ; Birkel & Hartmann (2020) ; Roscoe et al (2020) ; Kbah et al (2020) ; Shahbaz et al (2020) ; Baryannis et al (2019) ; Chowdhury et al (2019) ; Snoeck et al 2019 ). Sawik, T. (2019b) ; Mogos et al (2019) ; Stewart & Ivanov (2019) ; Gao et al (2019) ; Shahbaz et al (2019) ; Sawik (2019a) ; Bugert & Lasch (2018) ; Nakatani et al (2018) ; Kumar et al (2018) ; Ledwoch, et al (2018) ; Blackhurst et al (2018) ; Diabat et al (2019) .…”

Smart Manufacturing for Smart Cities—Overview, Insights, and Future Directions

“…The set of input features in this case could be similar to the ones aforementioned (process related) except that the target label would be binary (0 in case of normal operation and 1 in the case of machine breakdown or failure). Depending on the type of algorithm, i.e., either regression or classification, different types of metrics exist to gauge the accuracy of the model [35,36] and thereby improve the performance of the ML model either during cross validation or during subsequent training postiterative feature engineering.…”

“…Among the various algorithms evaluated, including LR, RR, lasso regression, multivariate adaptive regression (MARS), regression tree (RT), bagged RT, RF, boosted RF, SVM, k-NN, and ANN, it was observed that the RF algorithm gave the best prediction whereas ANN had the least predictive ability. Baryannis et al [36] developed and evaluated the risk prediction framework for supply chain risk management in an aerospace manufacturing supply chain. With details of a particular supplier for 36 677 product deliveries collected from the MES for a 6-year period, the authors used two binary classifiers, namely SVM and DT, to predict whether future deliveries of a particular supplier would be delayed or not, where the result indicated that the performance of the SVM classifier was better than DT in terms of higher precision and recall.…”

COVID-19 effect on supply and demand of essential commodities using unsupervised learning method

“…In recent years, both national and global level supply chain risk management attracted the attention of researchers and practitioners [1]. Big data and machine learning approaches help in the detection of emerging risks, maintenance of relevant reports, and initiate suitable actions for a reformation of the supply chain [1]. Using analytics, supply chain issues like track and trace, route optimization, Green Logistics can be resolved [10].…”

“…The challenging task faced by supply chain industries during a pandemic is predicting demand and supply, transportation issues, manpower issues, and government regulations. Managing these issues within and between the state has increased the attention of researchers towards the supply chain [1]. This type of disaster impacts mainly on customer behavior and preferences.…”