An ensemble technique to predict mental illness using data mining techniques

Abstract

The mental well-being of a person is their mental state. Chemical abnormalities in the brain cause mental health problems. It is important to monitor the mental health of different groups in order to predict health-related disorders. The community consists of working professionals and college students. It is widely believed that stress and grief affect people of all ages and backgrounds. Some serious mental health disorders, such as anxiety, bipolar disorder, and schizophrenia, often evolve and produce symptoms that can be recognized early. Such mental disorders could be avoided more successfully if abnormal mental states are detected in the early stages of the disease, allowing for additional care and treatment. This study analyzed the accuracy of four data mining techniques and introduced a new ensemble technique to improve their accuracy in identifying mental health issues. The data mining techniques are Logistic Regression, KNN Classifier, Decision Tree Classifier, and Random Forest. This paper provides scope for other researchers and practitioners seeking to achieve higher accuracy in identifying mental health issues using enhanced data mining algorithms to meet several accuracy criteria.