Prediction and Analysis of Employee Promotions Using Machine Learning

The primary objective of this study is to utilize classification techniques to develop predictive model algorithms that can determine employee eligibility for promotion and identify the most influential attributes affecting employee advancement. The dataset utilized in this research is sourced from Kaggle 2020 and comprises data from a multinational company, encompassing 54,808 rows and 13 columns. It covers nine departments across all age groups and includes records of actual promotions. Various predictive modeling techniques were employed, such as logistic regression, random forest, and ensemble models (Adaboosting and Gradient Boosting), to forecast employee promotions. The F1 score was chosen as the evaluation criterion for the models, with the final outcome indicating that random forest outperforms other classification algorithms. Furthermore, it was revealed that average training score, year of rating, and age are pivotal factors in predicting employee promotions; whereas departmental affiliation and gender do not exert any influence.