Leveraging Machine Learning Methods for Predicting Employee Turnover Within the Framework of Human Resources Analytics

Abstract

Employee turnover is a critical challenge for organizations, leading to significant costs and disruptions. This study aims to leverage Machine Learning (ML) techniques within the framework of Human Resources Analytics (HRA) to predict employee turnover effectively. The research evaluates and compares the performance of six widely used models: Decision Trees, Support Vector Machines (SVM), Logistic Regression, Random Forest, XGBoost, and Artificial Neural Networks. These models were implemented using the R programming language on an open-source dataset from IBM. The methodology involved data preprocessing, splitting into training, validation and testing sets, model training, and performance evaluation using metrics such as accuracy, sensitivity, specificity, precision, F1-score, and ROC-AUC. The results indicate that the Logistic Regression model outperformed the other models, achieving high accuracy and a good F1-score. The study concludes by emphasizing the importance of HRA and ML techniques in predicting and managing employee turnover, while discussing limitations such as class imbalance and the need for more rigorous performance evaluation. Future research directions include exploring alternative models, feature selection techniques, and addressing class imbalance.