Comparative Study of Employee Turnover Analysis Methods

Abstract

In the rapidly evolving landscape of human resource management, understanding and mitigating employee turnover has emerged as a paramount challenge. This research delves into a comparative analysis of various machine learning algorithms applied to the same dataset over several research papers, with the objective of identifying the most effective method for predicting employee turnover. Through a meticulous evaluation based on accuracy, F1 score, and ROC value, this study aspires to identify the best methods for predicting employee turnover.This study highlights the Artificial Neural Network (ANN) as the top algorithm for predicting employee turnover due to its outstanding performance across multiple metrics, while XGBoost and LGBM Regression are also identified to be highly effective, providing a good mix of efficiency, interpretability, and reliability.• Performance Metrics: The evaluation of algorithms is based on accuracy, F1 score, and ROC AUC score. While these metrics provide a robust framework for comparison, they might not capture all aspects of model performance relevant to turnover prediction, such as calibration and the cost-sensitive nature of false positives versus false negatives in real-world HR decision-making.