Identification of human resource analytics using machine learning algorithms

Abstract

Employee attrition is one of the most significant business issues in human resource (HR) analytics. This research aims to identify the most critical elements that contribute to employee attrition. Businesses operate heavily on employee training in order to maximize the returns they will offer to the company in the future. By utilizing the employee information value concept, it has been discovered that employee features such as overtime, the total number of projects and job level have a significant impact on attrition. To find the probability of new employee attrition, various classification algorithms such as decision trees (DT) classifier, logistic regression (LR), random forests (RF), and K-means clustering are used. A comparative analysis of the models with different rating scales is carried out for the highest accuracy. For prediction, four diverse machine learning (ML) algorithms such as LR, RF, DT classifier, and k-nearest neighbors (k-NN) are used. DT classifier outperforms with 97% of accuracy than other techniques. The effects of predictive ML techniques on the employee dataset show that RF evaluation outperforms other ML techniques followed by model of LR for the specific dataset if precision is the preferred metric. Identification of HR is forecasted using ML algorithms on employee data.