Factor-bridging algorithm for the prediction of job satisfaction: Developing country perspective

Abstract

Job satisfaction is crucial for both job seekers and employers. To ensure positive job satisfaction, companies must implement policies that consider employees’ perceptions of their work. This study utilizes machine learning and factor analysis to predict job satisfaction. Although, factor analysis has limitations, such as data quality, small sample size, and difficulty interpreting factors, machine learning algorithms can overcome these challenges. This study predicts job satisfaction using field data by combining factor analysis with machine learning algorithms. Factor loading values significantly impact classification algorithms such as Logistic Regression, Decision Trees, Support Vector Machines, and Random Forest. Especially, the Management Support, Equity, Non-Financial Compensation, and Financial Compensation feature variables are highly effective. They are used to predict job satisfaction with factor-1, factor-2, and factor-3 values. The Random Forest and Support Vector Machines algorithms have shown the importance of these values. The enhanced precision has been demonstrated visually to highlight the contrast when compared to the factor loading analysis and their corresponding eigenvalues.

Keywords

• Factor analysis
• Logistic regression
• Decision tree classifier
• Random forest
• Support vector machine
• Feature importance