Predicting Employee Job Satisfaction with Machine Learning

Abstract

Employee job satisfaction profoundly influences employee retention, productivity, and organizational success. Conventional statistical approaches often fail to capture the complex interplay of numerous factors influencing job satisfaction, whereas machine learning (ML) offers robust capabilities for analyzing multidimensional datasets. In the present study, we investigated the factors shaping employee job satisfaction using data, comprising 9,516 observations. Following dimensionality reduction, 33 key variables were identified and modeled to predict overall employee job satisfaction using Naïve Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), and Gradient Boosted Trees (GBT) models. Among these, the SVM and GBT models achieved the highest predictive accuracy of 0.99. The results highlight the relative importance of work-related and personal factors, providing actionable insights for human resource management to enhance employee retention and organizational success.