A global challenge continues to increase with the rise of mental health conditions, therefore a reliable and accessible screening tool to detect mental illness earlier in an individual's process of seeking professional services must be developed. Typical ways of diagnosing mental illness are subjective and may depend on self-report surveys or clinical assessments which may depend upon availability, stigma or inconsistencies when reporting symptoms. This paper describes a machine-learning classification system for mental health using structured survey data with; demographic, behavioural, psychological, and workplace indicators. The data set underwent systematic pre-processing through; data cleaning, numerical encoding, normalisation, and exploratory correlation before being used as a high-quality and interpretable input for creating the ML models. Supervised learning algorithms were used to train and evaluate each of the algorithms to determine predictive performance. The results indicate that ensemble models widely outperformed all of the other algorithms with respect to their resistance to noise and ability to identify complex feature inter-actions. All performance metrics such as accuracy, precision, recall, and F1-score confirm that the models were stable across data splits. Exploratory analysis also supported a coherent relationship between identifying psychological characteristics and the models developed from surveys to predict mental health. Overall, the proposed framework offers a practical, adaptable, and scalable approach to early mental-health screening, supporting data-driven decision-making in organizational and community well-being initiatives.
Keywords: Mental-health prediction, Random Forest classifier, Psychological assessment, Exploratory data analysis, Mental-health screening
How to cite this article: Patel K, Jani K, Short Title: Survey-Based Mental Health Prediction Mental-Health State Prediction Using Survey-Based Machine-Learning Techniques. Int J Drug Deliv Technol. 2026;16(4s): 1-8; DOI: 10.25258/ijddt.16.1-8