PREDICTIVE ANALYSIS OF STRESS LEVELS AMONG WOMEN EMPLOYEES IN SELF-FINANCING COLLEGES IN CHENNAI WITH BAYESIAN MODELS

This research paper attempt to investigates stress among women employees in selffinancing colleges in Chennai by analyzing socio demographic, behavioral, and psychological factors affecting their well-being. Data from 421 participants were preprocessed, scaled, and split into training (70%) and testing (30%) sets to evaluate three Bayesian algorithms: Gaussian Naive Bayes, Bernoulli Naive Bayes, and Bayesian Logistic Regression. Model performance was assessed using accuracy, confusion matrices, and classification reports, with Bayesian Logistic Regression demonstrating the highest accuracy of 85%. Feature significance was further analyzed using ANOVA, revealing that income and self-reported stress levels were the most influential factors. The results emphasize the importance of identifying key stressors and implementing targeted interventions to support mental health, productivity, and work-life balance for women employees in academic institutions.