HR Analytics and Employee Attrition Prediction Using Machine Learning

Abstract

The high rate of staff turnover in knowledge-based organizations is a key challenge. Invaluable tacit information, which is frequently the source of a company's competitive advantage, is often taken with departing personnel. If a company wants to keep a better competitive edge over its competitors, it should prioritize lowering staff churn. By forecasting attrition based on demographic and job-related factors, this study finds employee traits that aid in predicting employee turnover in organizations. With an IBM dataset of 14,999 samples and 10 features, up-sampling techniques and various machine learning algorithms are employed to find the best predictive model. Data visualization and analysis reveal significant factors and correlations. Further, the authors have used models to predict and analyze employee attrition and turnover. Using classifiers like k-nearest neighbors (KNN), support vector machine (SVM), decision trees (DT), and random forest classifiers (RF), four main tests were run on the IBM dataset to predict employee attrition.