Enhancing Employee Attrition Prediction Using Hybrid Machine Learning Models

Employee attrition prediction is crucial for organizations to mitigate the risk of losing top talent, ensuring smooth operations and reducing costs associated with recruitment and training. This project aims to enhance employee attrition prediction by utilizing hybrid machine learning models, combining both classical and advanced algorithms for better accuracy. The system integrates a Flask web application with a MySQL database for managing employee records, model training, and predictions. The system employs multiple machine learning models, including Stacking Classifier, Voting Classifier, XGBoost, Decision Tree, and Random Forest, to classify whether an employee is likely to leave the organization. Each model's performance metrics, including accuracy, precision, recall, and F1-score, are computed to evaluate and select the best-performing model. The system allows users to upload employee data in CSV format, and the prediction is made based on selected models. The web interface is user-friendly, providing clear feedback to users and making predictions accessible. The integration of machine learning models with real-world data through a web interface provides organizations with valuable insights into employee retention, helping HR departments to take proactive actions and reduce turnover.

Keywords: Employee Attrition, Machine Learning, Flask, MySQL, Hybrid Models, Prediction, Random Forest, XGBoost, Stacking Classifier, Voting Classifier, Decision Tree, Web Application.