Prediction of Workplace Absenteeism Time using Machine Learning

Abstract

Absenteeism in the workplace is a significant cause of lost productivity of the organization and the root cause of the company's performance to many employers. Managing absenteeism is inevitable, but making sudden changes without knowing the cause of the problem is a terrible mistake. This paper aims to develop a reliable workplace absenteeism prediction model using machine learning and natural language processing techniques to aid employers with analyzation of given minimal available information about the employees’ demographics. ‘Distance from residence to work,’ ‘disciplinary failure’ and ‘weight’ was negatively associated with absenteeism time in hours. ‘Age,’ ‘son,’ and ‘height’ were positively associated with absenteeism time in hours