Employee Promotion Prediction Using Improved AdaBoost Machine Learning Approach

Abstract

Employee promotion is an important aspect of the human resource management process. Due to different factors, it refers to the automatic improvement among the employees in an organization. Promoting employees from the lower level to the higher level brings a feeling of satisfaction among the employees. It improves their job satisfaction and motivation by providing more significant income, status, and responsibilities. By building up loyalty, promotion reduces employee attrition. Thus, it is difficult to accurately decide, whether an employee should or should not be promoted based on their current and past performance. So, human resource management does research about promotion, because there are a limited number of research about the finding of employee promotion prediction in the existing studies. First, to find the reasons for employee promotion, we need to analyze the research study for finding the factors which are related to the promotion. The aim of this research study is to implement an employee promotion prediction framework using machine learning. A modified AdaBoost classifier is used for automatic promotion prediction, and six machine learning techniques for instance, Support Vector Machine (SVM), Logistic Regression (LR), Artificial Neural Network (ANN), Random Forest (RF), XGBoost (XGB), and AdaBoost are applied in performance comparison. Through a complex assessment process, the performance of these supervised machine learning algorithms for predicting employee advancement is analyzed using assessment metrics on the employees' evaluation dataset for promotion prediction. The Artificial Neural Network (ANN) and AdaBoost model provide better results on this dataset than all traditional machine learning techniques. Finally, Our proposed modified AdaBoost approach outperformed all other methods evaluated with an accuracy of 95.30%.