Scholarlink

Employee attrition poses substantial challenges for organizations, disrupting stability and escalating costs associated with hiring and training. Using the IBM HR Analytics dataset with 1470 records and 30 features, the research explores demographic, job-related, and performance metrics such as age, gender, education, job role, salary, work life balance, and performance ratings. Predictive models like Decision Tree, Support Vector Machine (SVM), Logistic Regression, LSTM, Stacked LSTM, and Bidirectional LSTM were assessed using metrics like accuracy, precision, recall, and F1 score. Bidirectional LSTM emerged as the top-performing model, achieving accuracy (89.94%), recall (89.34%), precision (88.78%), and F1 score (89.43%), demonstrating its capability to effectively analyze patterns and predict employee attrition. The study helps organizations by identifying employees who might leave, allowing timely actions like promotions or support, highlighting key reasons for attrition, and enabling companies to create better retention strategies, which reduces turnover, improves employee satisfaction, and ensures smoother workforce planning.

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

Employee Attrition Prediction Using Ensemble Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Akshata Basapur,
  • Laxmi Metgud,
  • Sneha Tenginakai,
  • Tejas Kalghatgi,
  • Uday Kulkarni,
  • Shashank Hegde

摘要

Employee attrition poses substantial challenges for organizations, disrupting stability and escalating costs associated with hiring and training. Using the IBM HR Analytics dataset with 1470 records and 30 features, the research explores demographic, job-related, and performance metrics such as age, gender, education, job role, salary, work life balance, and performance ratings. Predictive models like Decision Tree, Support Vector Machine (SVM), Logistic Regression, LSTM, Stacked LSTM, and Bidirectional LSTM were assessed using metrics like accuracy, precision, recall, and F1 score. Bidirectional LSTM emerged as the top-performing model, achieving accuracy (89.94%), recall (89.34%), precision (88.78%), and F1 score (89.43%), demonstrating its capability to effectively analyze patterns and predict employee attrition. The study helps organizations by identifying employees who might leave, allowing timely actions like promotions or support, highlighting key reasons for attrition, and enabling companies to create better retention strategies, which reduces turnover, improves employee satisfaction, and ensures smoother workforce planning.