Data-Driven Attrition Prediction for Talentgigs, Chennai

Abstract

Staff turnover is a serious problem in any organization that impacts on organizational stability, productivity and competitiveness. High attrition results in extra recruitment and training expenses, lower employee morale and team cohesiveness. The traditional HR approach that employs reactive analysis and manual judgement are often found to be inadequate for forecasting. This work gives more emphasis on the understanding of the effectiveness of Data-driven machine learning approach to predict employee attrition in TalentGigs, Chennai. The goal of this study was to create and benchmark several machine learning models to determine which is the most accurate and reliable predictor of attrition risk. A synthetic employee attrition dataset was used as the secondary data source. The analysis tools employed were Python and libraries like Scikit-learn, Pandas, and XGBoost. The study result showed that the highest accuracy of 75.81% was obtained from the XGBoost algorithm during the prediction of Attrition. Overall, the study highlights the potential of AI-powered predictive models in HR, allowing organizations to make proactive, data-driven decisions that help reduce turnover, control costs, and create a more stable workforce.