Evaluation of energy sector providers for enhanced efficiency, performance, and sustainability using the digital, lean, agile, resilient, and green (DLARG) framework: a machine learning model

The energy sector plays a pivotal role in the modern industrial landscape, making the evaluation of raw material providers essential for enhancing efficiency, performance, and sustainability. This study introduces a comprehensive evaluation framework that incorporates key features from the digital, lean, agile, resilient, and green paradigms to foster viable business practices within the energy sector. Initially, we identify the primary criteria necessary for assessing raw material providers. Subsequently, we develop a machine learning–based model designed to determine the weights of these criteria and evaluate raw material providers’ performance. To achieve this, we employ the fuzzy best–worst method to compute the weights of the identified criteria. Performance assessment is conducted using a combination of data envelopment analysis and gradient boosting trees methodologies. We demonstrate the application of our machine learning–based model through a real-world case study in the energy sector. The results indicate that the most critical performance indicators include “Responsiveness,” “Robustness,” “Cost,” “Quality,” “Manufacturing Flexibility,” “Technical Capability,” “Restorative Capacity,” “Lead Time Flexibility,” and “Waste Management.” Furthermore, we validate the effectiveness of the fuzzy best–worst method by comparing its results with those from alternative methods, confirming its reliability. Similarly, the performance of the integrated data envelopment analysis and gradient boosting tree approach is benchmarked against traditional methods, demonstrating its validity and effectiveness. Overall, our findings provide valuable insights for supply chain managers, particularly within the energy sector, aiming to enhance their operational strategies and sustainability initiatives.