Non-Invasive Mental Health Prediction using Machine Learning: An Exploration of Algorithms and Accuracy

The potential of the machine learning in predicting mental health outcomes is investigated in this study. Two datasets were gathered: one of mental health patient questionnaires and the other of information from MRI scans of Alzheimer’s patients. The datasets were pre-processed using techniques such as stop word removal and lemmatization, and the processed data was encoded for increased prediction accuracy. To find the highest performing model, various algorithms such as Logistic Regression, Decision Tree, KNN (K-Nearest Neighbors), Adaboost, Random Forest, and Logistic Regression are examined. The findings indicated that machine learning algorithms can predict mental health outcomes with high accuracy, and that adding demographic, behavioural, and psychological factors can improve prediction accuracy even more. The study emphasizes the significance of creating accessible and accurate mental health prediction tools, as well as the promise of the machine learning in mental health evaluation.