Exploring the Accuracy of Machine Learning Models in Predicting Toxic Leadership: A Case Study in the IT Industry

This study delves into applying various machine learning models to predict toxic leadership in the Moroccan IT sector, employing Decision Trees, Random Forest, SVM, Neural Networks, Logistic Regression, and Ensemble Models (Voting Classifier). Evaluating key metrics, including precision, recall, f'l-score, and accuracy, the ensemble model emerges as the most robust predictor. Notably, the strongest contributors to toxic leadership, ranked in descending order, are Undermining Behavior, Narcissistic Traits, Unfear Treatment, and Fear of Retribution. This research aims to comprehensively understand toxic leadership prediction by employing a diverse set of models. Ensemble models showcasing superior predictive capabilities.