Journal of Management and Business Solutions 3092-7226 3 Serial Number 15 2025 10 10 Anticipating Leadership Competency Shifts: Supervised Learning Models for Executive Talent Futures Anticipating Leadership Competency Shifts: Supervised Learning Models for Executive Talent Futures 1 11 EN Journal Article 2025 07 23 <p>This study aimed to anticipate future shifts in executive leadership competencies by applying supervised learning models to predict emerging leadership capability patterns based on individual, experiential, and contextual predictors. The study employed a quantitative, predictive research design using supervised machine learning techniques. Data were collected from 312 senior and middle-level executives employed in public, private, and semi-governmental organizations in Tehran, Iran. Leadership competencies were measured through a validated multi-dimensional executive competency inventory capturing strategic thinking, adaptive decision-making, digital leadership, emotional intelligence, ethical leadership, innovation orientation, and systems thinking. Predictor variables included learning agility, exposure to digital transformation initiatives, frequency of strategic uncertainty, executive development experiences, and demographic characteristics. After data preprocessing and standardization, multiple supervised learning algorithms—multiple linear regression, support vector regression, random forest regression, and gradient boosting regression—were trained and evaluated using cross-validation procedures to ensure robustness and generalizability. Comparative model evaluation demonstrated that non-linear ensemble models significantly outperformed linear regression, with gradient boosting regression achieving the highest explained variance and lowest prediction error. Feature importance analysis revealed that learning agility, exposure to digital transformation projects, and strategic uncertainty frequency were the most influential predictors of anticipated leadership competency shifts, while traditional indicators such as tenure and educational level exhibited weaker predictive power. Cross-validation results indicated stable model performance, confirming the reliability of supervised learning approaches for forecasting leadership competency evolution. The findings indicate that executive leadership competencies evolve in predictable patterns shaped by learning-oriented and experiential factors, and that supervised learning models provide a powerful methodological tool for anticipating future leadership capability requirements and informing proactive talent development strategies.</p> https://journalmbs.com/index.php/jmbs/article/download/188/141