Developing a hybrid machine learning model for employee turnover prediction: Integrating LightGBM and genetic algorithms
Employee turnover poses a significant challenge to organizations, leading to increased costs and disruptions in workforce stability. In this study, we propose a novel hybrid model for employee turnover prediction that integrates Genetic Algorithms (GA) for feature selection with LightGBM for classif...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-06-01
|
| Series: | Journal of Open Innovation: Technology, Market and Complexity |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2199853125000927 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Employee turnover poses a significant challenge to organizations, leading to increased costs and disruptions in workforce stability. In this study, we propose a novel hybrid model for employee turnover prediction that integrates Genetic Algorithms (GA) for feature selection with LightGBM for classification. The GA component identifies the most relevant features, while LightGBM ensures high prediction accuracy and computational efficiency. The model was evaluated on two datasets—a public HR Analytics dataset and a proprietary organizational dataset—using multiple performance metrics including Accuracy, Precision, Recall, F1-Score, and AUC. The proposed GA + LightGBM model achieved an AUC of 0.94 and F1-Score of 0.87 on the HR dataset, and an AUC of 0.78 with F1-Score of 0.73 on the organizational dataset, outperforming conventional models such as Random Forest, Logistic Regression, and even standalone LightGBM. The results demonstrate that the proposed method offers a balanced and interpretable solution for turnover prediction, outperforming existing approaches in both predictive power and practical applicability. This work contributes to HR analytics by offering a transparent, scalable, and high-performing predictive framework that supports effective retention strategies. |
|---|---|
| ISSN: | 2199-8531 |