Identification of Anxiety and Depression Using DASS-21 Questionnaire and Machine Learning

Identification or psychologically associated emotional activities through machine learning techniques and artificial intelligence is found to be widely explored in various research publications. Studies revealed the relevance of machine learning techniques along with artificial intelligence for the recognition of human emotions such as sadness, anger, happiness, etc. using datasets like face image, person video, audio, questionnaire-response, etc. Human emotions come from psychological activities that may he affected by outside daily life routines. The proposed study reveals the configuration of anxiety and depression symptoms from the questionnaire-based dataset. In the present manuscript, we have used the standard DASS-21 questionnaire for the identification of anxiety and depression by applying machine learning algorithms on the user responses. We have analyzed and presented the comparative performance or five classification algorithms i.e., SVM, Decision Trees, Random Forest, Naïve Bayes and KNN on the aforementioned problem of identification of users under Depression and Anxiety.