A Hybrid AI-DEA Decision-Support Framework for Predictive Efficiency Analysis in Higher Education

Abstract

This study presents a novel framework integrating Artificial Intelligence (AI) with Data Envelopment Analysis (DEA) to enhance efficiency measurement, prediction, and decision-making in higher education institutions (HEIs). Addressing the static limitations of traditional DEA, the framework combines Machine Learning (ML), Deep Learning (DL), and optimization techniques to handle data uncertainty, select critical features, forecast future efficiency, and optimize resource allocation. Using simulated data for five hypothetical universities, representing diverse HEI archetypes, and extended to real-world datasets, the framework employs Fuzzy DEA and AI-driven models to achieve superior predictive accuracy and scalability, validated through efficiency scores (θ) and metrics (MSE, R², ROC-AUC). An AI-powered Decision Support System (DSS) with an interactive dashboard visualizes efficiency trends and provides actionable policy recommendations. This framework enables policymakers to allocate resources more effectively, enhancing HEI competitiveness and governance. It contributes to AI-driven operational research, offering scalable solutions for educational management and institutional benchmarking.Accordingly, this study is positioned as a methodological framework paper, focusing on the design and integration of an AI-enhanced DEA system for predictive and prescriptive efficiency analysis, rather than as a statistically generalizable assessment of HEI performance.