IJARSAI Int. J. Adv. Res. Sci. Anal. Infer.
Arya Soman
Research Scholar in Commerce
Government College
Nedumangad
University of Kerala
Thiruvananthapuram, Kerala, India
Email: aryasoman001@gmail.com
Rasheeda R.V.
Research Scholar in Commerce
Government College
Nedumangad
University of Kerala
Thiruvananthapuram, Kerala, India
Email:Phduokrasheeda@gmail.com
Rohit Krishnan M.
Allianz Services, India
Email: rohitkrishnanm@gmail.com
Dr. Rejani R. Nair
Assistant Professor
Government Arts College
Thycaud
University of Kerala
Thiruvananthapuram, Kerala, India
Email: rejanirnair@yahoo.co.in
Stability of the organisation and cost effectiveness depend on employee retention. Higher attrition rates cause significant financial and operational constraints as well as impede operations. This study looks at a machine learning approach for staff attrition forecasting, therefore providing businesses with useful information for preventative action. Data preprocessing, exploratory data analysis (EDA), and predictive modeling-all part of a methodical approach-are used in the paper. The dataset consists of necessary educational, professional, and demographic aspects. Managing missing values, encoding categorical variables, normalising numerical features, and correcting class imbalances using the Synthetic Minority Oversampling Technique (SMote) constitute part of the preprocessing processes. The exploratory study shows relationships between traits and target factors, suggesting that workers who put more hours of training are more likely to seek professional adjustments. With an accuracy of 85.97% combined with well-balanced precision and recall metrics, a comparative analysis of machine learning models-more especially, Logistic Regression, Random Forest, and Support Vector Machine (SVM)—showcases the Random Forest method attaining optimal performance. The study emphasises how well machine learning can help HR teams carry out targeted retention plans and advance organisational stability. This work improves predictive HR analytics by providing a thorough theoretical and methodological framework to address attrition problems and hence support dynamic workforce management solutions.