Machine Learning as a Data Science Tool to Predict Absenteeism for Factory Workers

Employee absenteeism results in the loss of billions of dollars in revenue for organizations around the globe. Studies estimate a 3% loss of scheduled labor hours due to these unexpected and uncertain absences. Hence predicting these unforeseen absences would allow efficient workforce planning and help lower overtime costs for organizations. This paper has looked at the issue from the POV of the management to explore the absenteeism in blue-collared workers employed at the Research and Development center of one of India’s biggest fast-moving consumer goods companies. The paper has used basic Machine Learning algorithms models such as Logistic Regression, XGBoost, Random Forest, and Support Vector Machines to predict absenteeism in these employees by analyzing the data collated from various employment records in the company over two years (2017-2019). The paper has also compared their accuracy in making these predictions, thereby letting us know if these algorithms would suffice in making these predictions or if there is a need to deploy more superior and powerful computational models to do so.