Predicting Mental Health Disorders in the Technical Workplace: A Study on Feature Selection and Classification Algorithms

Mental disorders are increasingly common among technical employees, posing significant challenges in the workplace due to high levels of stress. Recognizing and accurately predicting diverse situations is crucial for promoting mental health in healthcare settings. This study aims to predict mental health disorders using feature selection algorithms on the tech survey dataset, which consists of 61 features related to mental health attributes and frequency in the global technical workplace. Multiple machine learning classification algorithms were applied to the best features selected by RFECV, LASSO, and RFE. Performance metrics such as precision, accuracy, and recall were used to determine the optimal models. The results, discussed in an aggregated table, reveal the percentage of technical employees experiencing mental disorders. Our method performance in terms of classification accuracy has been proven to be much higher when compared with many competing feature selection strategies. The proposed research achieved a 79% accuracy rate by evaluating various classification algorithms alongside feature selection methods.