The Predictive Model of Mental Illness using Decision Tree and Random Forest classification in Machine Learning

Ministry of HFW, Government of India ordered the NIMNS - National Institute of Mental Health and Neuro Sciences, Bengaluru, in alliance with 15 institutions from across India and made a survey on mental health issues. This commission covered 12 states, one among that is Punjab from Northern region. As per the report, 15% of the adults in India need treatment for mental disorder. Machine Learning is one of the most substantial proportions of Artificial Intelligence. Machine Learning is widely used in many fields like online fraud detection, speech recognition, and social media. It plays a vital role in healthcare sector. This boosts the interest on the detection of the mental illness using machine learning algorithm. The big challenge is to predict the state of mind. Psychologists impose assessment and therapy to their patients by one to one physical interactions. There are multiple causes to put the person into critical situation like depression, pressure etc. Hence, this research paper proposes an ideal solution to identify the sickness in the person by checking with the recorded dataset. The most preferred Supervised Machine Learning algorithm, Decision Tree Classifier is used for this purpose. The initial goal of the Decision Tress is to create training ideal which is used to forecast the target variable class. The parameters considered here are anxiety disorder, depression disorder and the stress. Random Forest algorithm is applied to predict the illness in the people. The result obtained is to have accurate prediction level compared to the existing model.