Abstract
In recent years, the application of Machine Learning (ML) to predict mental disorders has gained significant attention due to its potential for early pre-diction. This study highlights the challenges of ML in mental disorders predic-tion, such as missing data in mental health datasets, by comparing four data im-putation methods: Mode, Multivariate Imputation by Chained Equations, Hot Deck, and K-Nearest Neighbor (K-NN) to enhance predictive accuracy; and uti-lizing four ML classifiers and three ensemble methods: Bagging, Boosting, and Stacking, with Mode and K-NN imputation datasets to show consistent perfor-mance. The study ultimately contributes to early mental disorder diagnosis and intervention in alignment with the United Nations Sustainable Development Goal 3 (SDG 3) for global health and well-being, by highlighting ML and data impu-tation’s potential in mental health analysis and paving the way for further ad-vancements in the field.