An Integrated Approach of MCDM Methods and Machine Learning Algorithms for Employees' Churn Prediction

Abstract

Employee churn is a notable nuisance for organizations to maintain a cost-effective position and brand strategy. This research has proposed an integrated approach to overcome this issue systematically. All employees in this study were divided into three categories using a combination of machine learning algorithms and Multi-Criteria Decision Making (MCDM) techniques. The techniques for Order of Preference by Similarity to Ideal Solution (TOPSIS) and the Analytical Hierarchy Process (AHP) have been consolidated. The dataset's feature importance was retrieved, and AHP was used to determine the criteria that caused the most turnover. TOPSIS was given the derived criterion weights from AHP to rank every personnel according to their propensity to depart the organization. To estimate staff turnover, seven machine learning algorithms were applied. After comparing the results, the Random Forest algorithm produces the best accuracy for assessing employee churn.