A Machine Learning-Based Employee Layoff Prediction System

Employee layoffs can significantly impact organizations and employees financially and psychologically. Predicting potential layoffs in advance can help companies implement proactive workforce management strategies. This paper proposes a ML-based Employee Layoff Prediction (MLELP) model that leverages Advanced Random Forests with Synthetic Minority Over-sampling Technique (SMOTE) for balancing the class and Recursive Feature Elimination (RFE) for optimal feature selection. The proposed system enhances prediction accuracy by integrating data pre-processing, feature engineering, and real-time API connectivity while ensuring fairness and transparency. The system is integrated into a webbased dashboard, where the frontend is developed using HTML and CSS, and the backend utilizes the Python Flask framework with a RESTful API for seamless interaction. The model considers various employee-related, financial, and organizational factors to generate layoff risk scores. The system provides HR professionals with an intuitive interface to visualize predictions, analyze contributing factors, and make data-driven decisions. Real-time data from HR databases, company financial records, economic markets, and employee surveys are used for evaluation. This proposed model predicts the lay-off likelihood with 99.5% accuracy.