Comparative Analysis of Entropy Weight Method and C5 Classifier for Predicting Employee Churn

Abstract

As per Aon, three out of four employees leave voluntarily, hence making employee churn (EC) a key issue that has an adverse impact on the productivity and growth of the organisation. Different methods have been used by decision makers to find the contribution of monetary and non-monetary factors but comparison of methods have been overlooked to find the best suited method. Researchers have identified the factors that lead to EC in the existing literature. However, it is important to know the degree of contribution of each factor. In this paper we have used two methods, multi-criteria decision-making method i.e., entropy weight method that derives relative weight for each factor/criterion. With the integration of advanced technology like machine learning (ML), various algorithms have been tested to predict EC. We have compared EWM and top-performing ML classifier out of four classifiers are C5, neural network (NN), decision tree (DT) and support vector machine (SVM) The novel contribution of our study is the comparison of two these methods and providing useful understanding for HR professionals in an organizational context.