Data Science Process Pipeline to Solve Employee Attenuation: Attrition prediction

Abstract

Industries are the backbone of the country visibility at global market. Employs and their capabilities are the back bone of industry growth. Therefore, most of the industries have started Human Resource Department to recruit and train employs for contributing company with huge investment. In other side, it become difficult to retain their employs because of death, retirement and resignation due to good offers from other companies. Therefore, it become more difficult to the HRD to transform their experience, knowledge and investment with new persons, it is widely recognized as Employ Attrition or Attenuation. In this scenario, HRD team is more responsible in identifying the reasons whether an employee can continue or leave company. This paper focus on data science process pipeline to identify the factors for employ attrition and predict whether an employ leaves the company or not. The proposed model is integrated with the various prediction models and returns the predicted value which is the best. In addition to that it is also investigated with Top K-best features selected from chai square test analysis and train the model to get accurate prediction value. The experimental analysis shows with accuracy values of the prediction methods for the given dataset