Enhancing Supply Chain Management - A Comparative Study of Machine Learning Techniques with Cost-Accuracy and ESG-Based Evaluation for Forecasting and Risk Mitigation

Aug 3, 2026

In today’s volatile market environment, supply chain management (SCM) must address complex challenges such as fluctuating demand, fraud, and delivery delays. This study applies machine learning techniques—XGBoost and recurrent neural networks (RNNs)—to optimize demand forecasting, inventory policies, and risk mitigation within a unified framework. XGBoost attains high forecasting accuracy (MAE = 0.1571, MAPE = 0.48%), while RNNs excel in fraud detection and late delivery prediction (F1-score ≈ 98%). The work adds Cost–Accuracy Efficiency (CAE) and CAE-ESG metrics to select models based on performance, cost-efficiency, and ESG alignment, supported by SHAP-based interpretability and business impact via improved CLV and reduced churn.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages45

Reading Time1 h 53 min

AccessFree to download

Introduction

Gaps in current ML-driven SCM research

Abstract

Objectives and methods

Evaluation beyond accuracy

Metrics, interpretability, and business impact

Article Enhancing Supply Chain Management: A Comparative Study of Machine Learning Techniques with Cost–Accuracy and ESG-Based Evaluation for Forecasting and Risk Mitigation Mian Usman Sattar 1, *, Vishal Dattana 2, Raza Hasan 3, *, Salman Mahmood 4, Hamza Wazir Khan 5 and Saqib Hussain 6 Academic Editors: V. Ravi., Suresh Subramoniam and Bijulal D. Received: 30 May 2025 Revised: 17 June 2025 Accepted: 20 June 2025 Published: 23 June 2025 Citation: Sattar, M.U.; Dattana, V.; Hasan, R.; Mahmood, S.; Khan, H.W.; Hussain, S. Enhancing Supply Chain Management: A Comparative Study of Machine Learning Techniques with Cost–Accuracy and ESG-Based Evaluation for Forecasting and Risk Mitigation. Sustainability 2025, 17, 5772. [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)su17135772 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://creativecommons.org/](https://creativecommons.org/)[ ](https://creativecommons.org/)[licenses/by/4.0/](licenses/by/4.0/)) . 1 College of Science and Engineering, University of Derby, Kedleston Road, Derby DE22 1GB, UK 2 Department of Computer Science and Management Information System, Oman College of Management & Technology, P.O. Box 680, Barka 320, Oman; [vdattana@ocmt.edu.om](vdattana@ocmt.edu.om) 3 Department of Science and Engineering, Southampton Solent University, Southampton SO14 0YN, UK 4 Department of Computer Science, Nazeer Hussain University, ST-2, Near Karimabad, Karachi 75950, Pakistan; [salman.mahmood@nhu.edu.pk](salman.mahmood@nhu.edu.pk) 5 Department of Business Studies, Namal University, Mianwali 42250, Pakistan; [hamza.wazir@namal.edu.pk](hamza.wazir@namal.edu.pk) 6 Department of Computer and Information Sciences, Northumbria University, Newcastle upon Tyne NE1 8QH, UK; [saqib2.hussain@northumbria.ac.uk](saqib2.hussain@northumbria.ac.uk) * [Correspondence: u.sattar@derby.ac.uk](Correspondence: u.sattar@derby.ac.uk) (M.U.S.); [raza.hasan@solent.ac.uk](raza.hasan@solent.ac.uk) (R.H.) Abstract In today’s volatile market environment, supply chain management (SCM) must address complex challenges such as fluctuating demand, fraud, and delivery delays. This study applies machine learning techniques—Extreme Gradient Boosting (XGBoost) and Recurrent Neural Networks (RNNs)—to optimize demand forecasting, inventory policies, and risk mitigation within a unified framework. XGBoost achieves high forecasting accuracy (MAE = 0.1571, MAPE = 0.48%), while RNNs excel at fraud detection and late delivery prediction (F1-score ≈ 98%) . To evaluate models beyond accuracy, we introduce two novel metrics: Cost–Accuracy Efficiency (CAE) and CAE-ESG, which combine predictive performance with cost-efficiency and ESG alignment. These holistic measures support sustainable modelselection aligned with the ISO 14001, GRI, and SASB benchmarks; they also demonstrate that, despite lower accuracy, Random Forest achieves the highest CAE-ESG score due to its low complexity and strong ESG profile. We also apply SHAP analysis to improve model interpretability and demonstrate business impact through enhanced Customer Lifetime Value (CLV) and reduced churn. This research offers a practical, interpretable, and sustainability-aware ML framework for supply chains, enabling more resilient, costeffective, and responsible decision-making. Keywords: demand forecasting; inventory optimization; machine learning; XGBoost; RNNs; risk mitigation; supply chain management; model interpretability; Cost–Accuracy Efficiency (CAE); ESG metrics; sustainable supply chains 1. Introduction In the contemporary landscape of global business operations, SCM has become increasingly complex, requiring advanced analytical methodologies to address its multifaceted challenges. Traditional SCM practices, which often rely on hi

FAQs

Which machine learning models are used for SCM optimization and risk mitigation?

The study uses XGBoost for demand forecasting and recurrent neural networks (RNNs) for fraud detection and late delivery prediction within a unified framework.

How does the paper evaluate models beyond traditional accuracy measures?

It introduces two metrics, Cost–Accuracy Efficiency (CAE) and CAE-ESG, combining predictive performance with cost-efficiency and ESG alignment to guide sustainable model selection.

What evidence of business impact and interpretability does the research provide?

SHAP analysis is applied to improve model interpretability, and results indicate enhanced Customer Lifetime Value (CLV) and reduced churn, supporting tangible business outcomes.

Aug 3, 2026

Upload by Seraphina