Review on Depression Detection Using Machine Learning Techniques

Machine learning has achieved notable advancements in the medical field by improving the accuracy, precision, and analysis of diagnostics while also reducing labour-intensive tasks. With strong evidence, machine learning has exhibited the capability of detecting mental distress including depression. It is fact that a large proportion of the population is currently experiencing depression, which is considered to be a global burden of disease. Hence there is a critical need to develop an early depression detection model. This article aims to improve our understanding the present state of depression prediction models by exploring issues and identifying difficulties. This review relies on the examination of audio, video, images, PHQ questions, Social media (Twitter and Facebook), EEG signal depression detection models using different machine learning approaches. This paper reviews indicators of depression and provides a summary evaluation of diverse research studies encompassing their performance parameters.