Mental Health Prediction using Machine Learning Models and Large Language Model

Mental health is integral to overall well-being, impacting human ability to deal with challenges in life. Machine learning and AI hold promise in predicting mental illnesses by analysing behavioural patterns, aiding in early detection and intervention. This proactive approach can mitigate symptom escalation, improving mental health outcomes. This research study introduces a novel predictive framework integrating ensemble learning techniques and large language models (LLMs). Initially, ensemble learning, including AdaBoost, voting, and bagging, constructs a robust model with Random Forest emerging as optimal. Subsequently, a Large Language Model (LLM) enhances the pipeline. User input triggers mental health prediction by Random Forest, forwarded to a Google Gemini model via an API key, generating personalised insights, marking a significant advancement in mental health prediction.