Optimizing Supply Chain and ERP Integration for Manufacturing Productivity Using Model Driven Extreme Learning Machine Models

Modern businesses use performance metrics to monitor, assess, and evaluate operations, especially in complex systems like ERP-based supply chains. Quantitative methods to identify and fix operational inefficiencies grow more crucial as these technologies spread. This study presents a total-related cost-based method for assessing a three-tier ERP-based supply chain. Preprocessing includes data normalisation, training/testing split, and bias reduction; feature selection uses penalised linear regression, principal component analysis, and correlation; and model training uses Model Driven Extreme Learning Machine (ELM).The Model Driven ELM outperformed GRU and BERT in accuracy and reliability with 93.59% accuracy. This comprehensive approach improves ERP-based supply chain performance assessment, operational decision-making, and cost-efficient optimisation in complex business systems.