Prediction of Stress Level on Indian Working Professionals Using Machine Learning

Introduction

Modern society is witnessing a continuous deterioration in the occupational health of working professionals in India. According to the latest reports for the year 2020, by ADP Research Institute, 70 percent of the Indian workforce is suffering from stress at their workplace which is a matter of grave concern. Survey reports by 1to1help, an Employee Assistance Program provider in India have found that there has been a large increase in the number of employees who are highly depressed or are vulnerable to suicides due to an increase in stress in their lives. Another surveyor, Optum affiliated with Nasscom, observed that almost half of the Indian employees suffered from some type of stress or mental health issue. They surveyed eight lacs employees in seventy major organizations, each with a minimum staff of four thousand and five hundred. The results were alarming and stated that the ratio of employees at high risk of taking their lives due to stress had increased from 4% to 8% in two years. They also found that family, money, and job were the most common stressors amongst the employees.

Parenting, pregnancy, change, caregiving, and social isolation were some more factors that led to stress. Some employees also fear for their job security in the restructuring of the organization even though they are better paid. It is specifically true for managers who are in a middle-level position and have the responsibility of kids as well as loans. Employees are also found to be struggling with stress in their personal lives due to broken relationships or marriages. Newer factors like social media also contribute to the intense peer pressure that expects people to have a certain lifestyle or be a certain image to qualify as successful. This undue and unmanaged stress leads to mental health issues viz. anxiety and depression, and damage to physical health by causing frequent headaches, elevated blood pressure, chest pain, upset stomach, and problems sleeping.

All these facts and figures provide enough motivation to find out the most influential factors that are the major contributors to stress amongst Indian working professionals and more importantly find the ranking of these factors so that one can work towards managing the most important stress contributors instead of all factors in general. In past, many authors have applied various statistical approaches to find the relation between different stress-causing factors and done critical analysis of the same.

Our proposed methodology is based on a machine-learning algorithm for the prediction of the stress level of employees in advance and hence mitigating the risks associated with undiagnosed stress levels that may trigger serious health complications. The most distinguishing feature of this piece of research is the extraction of factors in order of importance as contributing factors that cause stress in Indian working professionals. A Graphical User Interface has been developed by using Python tkinter for the prediction of stress levels. Thereafter, the strategies to reduce stress can be applied to make the lives of working professionals comfortable. This study utilizes some useful tips from Misra, 2020 that have elaborated the complete procedure from selection of the topic of research to writing conclusion section in a very systematic manner.

The paper comprises of following sections: Background, methodology for research (data preprocessing, exploratory data analysis, development of predictive model using machine learning algorithm, extraction of important stress-causing features using machine learning, GUI designing), conclusions, limitations of the study, and finally recommendations and future scope.