Using inverse DEA and machine learning algorithms to evaluate and predict suppliers’ performance in the apple supply chain

Abstract

In an era where sustainability is paramount, this study presents a novel analytical framework that uses inverse data envelopment analysis (IDEA) and machine learning to optimize supplier performance in Apple's supply chain. We introduce the application of IDEA to recalibrate operational benchmarks, focusing on substantial CO2 reduction, and couple this with advanced predictive algorithms to proactively steer supplier practices toward Apple's sustainability targets. The study systematically simulates scenarios to achieve 30%–50% CO2 cuts, setting a precedent for environmental strategy integration. Beyond environmental metrics, our approach rigorously evaluates economic indicators such as earnings persistence and market recognition, providing a holistic view of supplier viability. The research employs machine learning, notably random forest and k-Nearest neighbors, to distill unbiased insights from historical data, ensuring precise, objective supplier assessments. Our findings offer a strategic blueprint for corporations and supply chain stakeholders aiming to embed sustainability into their operational ethos, supported by data-centric decision-making. These methodologies and insights contribute valuable perspectives to sustainable supply chain management and predictive analytics discourse. Finally, the study offers guidance on implementing CO2 reduction strategies and evaluating suppliers' economic performance, aligning with Sustainable Development Goals (SDGs) objectives.