A Robust Data-Driven Framework for Evaluating National Industry 4.0 Performance: A Hybrid Learning and Decision-Making Approach

Abstract

 This study aims to suppress This study proposes a robust, data-driven framework to evaluate national Industry 4.0 performance by integrating unsupervised clustering, neural network– based importance estimation, permutation feature importance, and multi-criteria decision-making. Unlike conventional approaches relying on predefined or equal weights, the proposed framework derives criterion weights directly from data while accounting for structural uncertainty across alternative clustering schemes. The resulting robust weights are incorporated into TOPSIS to generate consistent country rankings. An empirical application using the Network Readiness Index, Global Innovation Index, and National Cyber Security Index demonstrates the effectiveness of the framework. The results identify human capital, governance capacity, and cybersecurity preparedness as the most influential dimensions of Industry 4.0 performance. Sensitivity analysis and equal-weight comparisons further confirm the stability of the rankings, providing a transparent and reliable framework for supporting national digital transformation strategies.