Scholarlink

Employee attrition, where employees leave the company, presents significant challenges for businesses globally, leading to increased costs, talent loss, and operational disruptions. Addressing the root causes of attrition is essential for maintaining organizational stability and productivity. This research focused on addressing the predictive challenge of employee attrition through the development of a novel Deep Neural Network (DNN)-based model. To enhance the model’s interpretability, eXplainable Artificial Intelligence (XAI) techniques such as SHapley Additive exPlanations (SHAP), and Local Interpretable Model-agnostic Explanations (LIME) were integrated. The DNN model achieved an accuracy of 95.71%, with precision, recall, specificity, F1-score, and ROC AUC score all exceeding 0.95. These results demonstrate the model’s effectiveness in identifying employees likely to leave. Furthermore, the insights provided by XAI techniques, both globally and locally, verified well with established human resource management knowledge, offering valuable perspectives on the factors driving employee attrition.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating Explainable AI with Deep Neural Networks to Improve Employee Attrition Prediction

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • G. P. Sachin Lakshan,
  • Nethmi Wijesinghe

摘要

Employee attrition, where employees leave the company, presents significant challenges for businesses globally, leading to increased costs, talent loss, and operational disruptions. Addressing the root causes of attrition is essential for maintaining organizational stability and productivity. This research focused on addressing the predictive challenge of employee attrition through the development of a novel Deep Neural Network (DNN)-based model. To enhance the model’s interpretability, eXplainable Artificial Intelligence (XAI) techniques such as SHapley Additive exPlanations (SHAP), and Local Interpretable Model-agnostic Explanations (LIME) were integrated. The DNN model achieved an accuracy of 95.71%, with precision, recall, specificity, F1-score, and ROC AUC score all exceeding 0.95. These results demonstrate the model’s effectiveness in identifying employees likely to leave. Furthermore, the insights provided by XAI techniques, both globally and locally, verified well with established human resource management knowledge, offering valuable perspectives on the factors driving employee attrition.