Mental Health Treatment Prediction Using Machine Learning

The main focus of this work is to identify and predict whether individuals experiencing a mental disorder have pursued treatment. This has been examined, trained, and tested using a survey dataset of common people with different age groups, gender, and working status. Here we combine predictive analytics, classification, and statistical summaries of patients who have undergone treatment for mental illness using various machine learning algorithms and predictive models. Mental illness does not have a single cause. Several factors contribute to mental illness, such major challenging factors considered in this work are employment status, age and gender. Patients those who have informed about mental health disorder, diagnostic models separating BD - Bipolar Disorder against MDD - Mental Depressive Disorder are trained and validated using machine learning algorithm named extreme gradient boosting with nested cross-validation. Core predictors included elevated treatment taken or not, age, and gender. Additional validation in participants with family history, work interface, interviews attended and so many other prediction factors.