Identifying Patterns and Predicting Employee Turnover Using Machine Learning Approaches
Aham Edward Kanuto,
DOI: https://doi.org/10.58970/ijsb.2373
2024-01-01
Abstract:Employee turnover poses significant challenges for organizations, impacting productivity, morale, and financial stability. Identifying patterns and predicting employee turnover using machine learning approaches can help organizations proactively address retention issues and optimize workforce management strategies. The current study analyzed a dataset comprising 4653 valid respondent records sourced from Kaggle, containing diverse attributes related to employees' educational backgrounds, work history, demographics, and employment-related factors. Through exploratory data analysis and feature selection, the study identifies key predictors of employee turnover, including factors such as education, joining year, city, payment tier, age, gender, ever benched status, and experience in the current domain. The researcher employs three machine learning algorithms—K-Nearest Neighbors (KNN), Decision Tree, and Support Vector Machine (SVM)—to predict employee turnover based on these factors. Evaluation metrics such as accuracy, precision, recall, and F1-score were utilized to assess the performance of each model. Additionally, techniques such as the Synthetic Minority Over-sampling Technique (SMOTE) were applied to handle class imbalance in the dataset. The findings reveal distinct characteristics and performance of each model, with the Decision Tree model exhibiting the highest accuracy and predictive capability. Through comprehensive analysis and model evaluation, this study contributes valuable insights into employee turnover prediction, enabling organizations to develop targeted retention strategies and foster a more engaged and stable workforce.