Automated Early Prediction of Employee Attrition in Industry Using Machine Learning Algorithms

Abstract

Several companies and organizations all over the world face redoubtable challenges in recruiting eligible and talented employee candidates, while at the same time enduring the risk of employee attrition caused as a result of death, resignation, or retirement. Employee attrition is one of the key issues that firms face today when endeavoring to retain their valuable and skilled workers. Losing talented employees results in a negative impact on the company's performance. Organizations suffer a huge loss in terms of recruitment costs, training costs, etc. which are incurred for an employee's upliftment. The departure of such employees can also reduce morale, increase stress, and over-burden the team members. Businesses can no longer rely on conventional business practices to promote growth due to the increase of a massive amount of data and the continuously changing preferences. Hence, artificial intelligence and machine learning open up a new realm of possibilities to drive business growth through actionable insights. The objective of this study is to determine why employees depart on their initiative, what might have kept them from doing so, and how to use this information to predict attrition risk. This research study uses three machine learning models to conduct the employee attrition prediction on the IBM Watson dataset having 35 features. The results are described in terms of performance metrics, and the Logistic Regression machine learning method, with an accuracy of 87% and the best recall rate (0.36), generated the best results for the dataset.

Ähnliche Arbeiten