Prediction of general anxiety disorder using machine learning techniques

Abstract

Today, the increase in mental health problems, the variable nature of mental health and the lack of sufficient number of mental health professionals have led to the search for machine learning that applied to mental health problems extensively, and its use in the field of health is considered as a new hope. Mental disorders are a health illness that affects a person's emotions, reasoning, and social interaction. Early diagnosis and the application of the right treatment after the correct diagnosis have always been the expectation of all humanity. As technologies develop, machine learning has started to attract attention in the field of medicine with the development of diagnostic methods. The aim of this study is to conduct classification studies by using machine learning methods in the diagnosis process of anxiety disorder diseases. A publicly available dataset of 672 people's Generalized Anxiety Disorder 7-item (GAD-7) responses during the COVID-19 period is used. This study demonstrates that it is possible to classify mental health status with 0.97 accuracy rates with the Support Vector Machine algorithm, which has a higher performance than other algorithms.