Artificial Intelligence Applications in Educational Management: Improving Decision Quality and Organizational Performance

Innovative approaches to the use of artificial intelligence (AI) in educational institutions are progressively changing administrative decision-making from retrospective reporting to predictive, explainable and evidence-guided action. We propose an Artificial Intelligence Educational Management for Decision Quality and Organizational Performance (AIEM-DQOP) framework for improving student success analytics, dropout-risk detection, faculty & course analytics, resource allocation, as well as institutional dashboarding and governance accountability. The study is based on the Open University Learning Analytics Dataset (OULAD), which acts as a primary benchmark dataset, with the UCI Higher Education Students Performance Evaluation dataset used for external validation. The methodology consists of a PRISMA-style review, data preprocessing and feature engineering, the construction of machine-learning and deep learning models, explainable artificial intelligence and fairness audit paired with human-in-the-loop decision control. Comparative models are logistic regression, random forest, XGBoost, LSTM and a hybrid XAI-assisted ensemble. The predictive performance, decision-quality indicators and organizational performance indicators are used to evaluate the framework. The results demonstrate that while hybrid provides better benchmark performances relative to baseline methods, governance improves interpretability, accountability, and decision tracing. A Q2 Scopus-level educational technology and higher-education management research contribution is an integrated, reproducible and governance-oriented model.