Employee Attrition: Prediction, Analysis Of Contributory Factors And Recommendations For Employee Retention

In today's highly competitive and demanding work environment, gaining accurate insights into what causes employee attrition will enable organizations to work on improving factors influencing it, so that they can retain excellent and hardworking employees — and by extension, continue to maintain the quality of the products or services they deliver. Moreover, such insights will also help employees to be better informed about the nature of the organization they intend to join. Based on the IBM HR Analytics Employee Attrition & Performance dataset, we have designed a system which assesses the importance of each feature that contributes towards each employee's possible attrition from the company. Through data visualization, feature importance calculation using SHAP (SHapley Additive exPlanations) index and prediction of attrition using machine learning models, we have determined which factors contribute the most towards employee attrition. A recommendation system is implemented based on the user-based collaborative filtering technique which churns out recommendations for what can be done to retain each employee based on their identified cause of attrition. This combination of computational methods can not only help employers to predict the possibility of attrition of new employees, but also prevent it.