Employee Attrition Prediction through Machine Learning: Advancing Human Resource Analytics and Workforce Management

Abstract

The term "employee attrition" describes a certain reduction in a company's workforce due to a wide variety of factors. Employee attrition may also be caused by various factors in companies. Nevertheless, the human resource managers have to be able to identify the signs of employee attraction at the initial stages. Organizational losses may be caused by employee attrition due to numerous reasons that include disruption of work and activities required to be done, re-employment and re-training cost, and loss of information. This study examines the accuracy of machine learning and deep learning models in forecasting employee turnover using the IBM HR Employee Attrition dataset. A number of models, including CNNs, were trained and evaluated against other methods, including DT, SVM, and LR. After comparing CNN model to more conventional machine learning algorithms, the experimental findings show that CNN model outperformed them with respect to accuracy (ACC) (92), precision (PRE) (96.67), recall (REC) (87), and F1-score (F1) (91.58). These findings indicate the efficiency of CNN to identify multifaceted trends in staff data, and it is a prospective HR analytics tool. These results demonstrate that predictive models based on deep learning have the potential to deliver actionable information to an organization to lower attrition and reinforce strategic human resource management.