Machine Learning-Driven Employee Layoff Prediction Using Social Network Analysis

Bobbala Harshitha, Balika Mahesh

Managing workforce stability during organizational changes is a critical challenge for modern enterprises. This study proposes an intelligent prediction system to identify employees who are at potential risk of layoffs by analysing historical employee data and workplace interaction patterns. The model integrates multiple factors, including demographic details, job history, performance evaluations, salary information, and career growth indicators, to build a comprehensive employee profile. In addition, Social Network Analysis (SNA) techniques are employed to capture employee connectivity and influence within the organization using metrics such as degree centrality, betweenness centrality, and closeness centrality. After preprocessing and selecting relevant features, a Random Forest classifier is developed to detect patterns associated with layoff events. The model’s performance is evaluated using standard metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. Experimental results indicate that the inclusion of network-based features improves the model’s ability to identify at-risk employees more effectively compared to traditional approaches. The proposed system provides valuable insights that can support organizations in proactive decision-making, enabling timely interventions, better workforce planning, and improved employee retention strategies.