Support Vector Machine Based Prediction of Work-Life Balance Among Women in Information Technology Organizations

Supervised machine learning (ML) is used in this article to create a prediction model for the degree of work-life balance (on a scale of 1–5) of information technology (IT) women professionals. In total, 425 women working in the IT industry participated in this article. Data collected is based on a questionnaire with 40 parameters, which is the input for supervised ML. Different ML models, such as regression trees and support vector machines (SVM), are created and compared. The performance of ML models is evaluated using the coefficient of determination (
R
²
), mean absolute error, and mean square error. An optimized SVM model is found to have the highest accuracy for work-life balance prediction. Furthermore, a comparison between the prediction capabilities of an SVM model and a multiple regression model is done. A comparison of the two prediction models finds that the accuracy of SVM is higher than that of multiple regression. This article provides SVM as an alternative prediction modeling technique to multiple regression in organizational behavior studies. ML modeling, optimization, training, validation, and interpretation of results are done stepwise to act as a practical guideline for researchers interested in ML modeling.