Machine learning and deep learning techniques for effective supply chain management

Abstract

The growth in data volume and complexity of decision-making create the need for new data analytics methods, such as machine learning (ML) and deep learning (DL), which are critical for improving various tasks in supply chain management (SCM). Recent studies have highlighted the benefits of integrating DL/ML in SCM. These efforts have been applied to measures, such as accuracy, precision, sensitivity, recall, and F1 score which can be used to evaluate the results and outputs of the ML/DL techniques; however, they cannot reflect stakeholders’ viewpoints on their benefits and effects on organizations. Therefore, this work aims to introduce a Single-Valued Trapezoidal Neutrosophic Number Weighted Arithmetic Average (SVTNNWAA)-based method for assessing the effect of integrating ML/DL with SCM tasks to benefit organizations. The proposed method integrates the full consistency method (FUCOM) to provide consistent weights and a robust tool, with fewer comparisons. The proposed method is applied with a practical case study, and the results show that it can enable the decision-makers to assess and visualize the benefit rates of ML/DL which indicate the most suitable method for more effective SCM. Additionally, the results indicate that the proposed method can evaluate the effect of ML/DL in SCM from stakeholders’ viewpoints towards their benefits under neutrosophic environment which can handle uncertainty and indeterminacy.