No Detail Found!

Mental health treatment prediction for tech employee with the implementation of ensemble methods

Author(s)
MICHAEL HENRY Computer Science Department, BINUS Graduate Program – Master of Computer Science Bina Nusantara University, Jakarta, Indonesia
SANI MUHAMAD ISA Computer Science Department, BINUS Graduate Program – Master of Computer Science Bina Nusantara University, Jakarta, Indonesia
Abstract
Individual has two awareness, one of them is mental health. It is the ability of a person to recognize the pressure in work and interaction within social life. Statistics shown an enormous amount of 2.570 billion people in the world experienced mental health disorders including depression, anxiety, alcohol abuse, etc. By considering this fact, mental health is one of the most important life essentials which needs to be taken care of. This study proposes an implementation of ensemble methods such as Bagging, Light Gradient Boosting Machine (GBM) and Stacking with Binary Particle Swarm Optimization (BPSO) as a feature selection and Open Sourcing Mental Illness (OSMI) as the dataset for the purpose of predicting whether IT employees need a mental health treatment or not. This study shows that ensemble methods do not always give better prediction with Naïve Bayes and BPSO which has 88.44% accuracy and Stacking with Naïve Bayes as the meta classifier has 87.86% accuracy, a 0.58% accuracy difference. Based on the performance, this study also shows the best features for predicting mental health treatment needs where IT Employees who have the problems or the best features, need to consult a treatment.
Publication Details
Page(s) 2675-2685
DOI DOI not available
Published Journal: Journal of Theoretical and Applied Information Technology, Volume: 100, Issue: 8, Year: 2022

Request the full text from PASTIC

Keywords
machine learning, Mental health, Particle Swarm Optimization, Ensemble Method, Feature Importance
0

Citations

0

Downloads

298

Views