A Strategic AI Framework Integrating Optimisation and Game Theory for Emotional Competency Evaluation in Educational Leadership

Emotional intelligence (EI) is increasingly recognised as a critical determinant of effective school leadership; however, the direct and cascading effects of leaders’ EI on teachers and students remain underexplored across diverse global contexts. Empirical evidence indicates that empathy, self-awareness, relationship management, and social intelligence are core EI competencies, and leaders with higher EI positively influence teachers’ emotional re-framing, collective efficacy, and instructional practices. In parallel, artificial intelligence (AI), combined with game-theoretic decision modelling, has emerged as a powerful tool for enhancing academic performance through personalised and adaptive learning environments, while also supporting the development of EI. Motivated by these insights, this study proposes a Deep Belief Network with Feature Optimisation for Assessing Emotional Competence in Leaders (DBNFO-AECL) framework. The model applies min–max normalisation, correlation analysis, and Principal Component Analysis (PCA) for optimal feature selection, followed by Deep Belief Network (DBN) classification enhanced using Northern Goshawk Optimisation (NGO). Experimental results on benchmark datasets demonstrate superior performance over existing methods across multiple evaluation metrics.