A Quantitative Analysis of AI and Machine Learning Applications for Supply Chain Optimization

Abstract

This study examined Abstract 9,994 retail orders observed between 2014 and 2017 through Artificial Intelligence (AI) and Machine Learning (ML) approaches. Quantitative analysis methods were used in the research to investigate shipping efficiency for discrete net lots across modes to predict delivery times and discover profitability patterns. The delivery performance showed significant variations between shipping modes (M = 34.6 days, SD = 55.1), and standard shipping is served by most orders (59.7%) through longer delivery times. The analysis showed that sales averaged $230.00 (SD = $623.00), but profit margins were quite variable (M = $28.70, SD = $234.00), permitting optimization. In particular, machine learning models predicted delivery times, and return behavior were analyzed, and their relationship with discount rates (M = 16 SD = 21) was examined. Local optimization strategies were suggested as necessary based on identified regional shipping efficiency and profitability variations. This study contributes to understanding AI and ML applications for supply chain management by empirical evidence of their effectiveness in improving delivery performance and profitability and what aspects of operating these applications could be accelerated.