Adaptive Neuro Fuzzy-Based Depression Detection Model for Students in Tertiary Education

Depression is a severe mental disorder with characteristic symptoms such as sadness, feeling of emptiness, anger, anxiety and sleep disturbance as well as general loss of initiative and interest in activities. The effects of late diagnosis of depression have culminated in many students dropping out of school, becoming unfulfilled in life and thereby posing a threat to human and environmental stability. Linear models for depression detection lack the intelligence to decode the non-linear interactions and imprecision proliferated in depression data. This work is aimed at utilizing Adaptive Neuro Fuzzy Inference (ANFIS) model furnished with intelligence for handling imprecision and non-linear modelling for diagnosis of depression. ANFIS model for depression detection was designed to provide a means of handling imprecision that characterizes depression detection attributes. Data collected from questionnaires administered to students in University of Uyo were analyzed. MATLAB programming tools were deployed for implementation of the model. The proposed fuzzy model identified and classified depression cases with 94.21% accuracy. The system would assist in early detection of depression in students studying in tertiary institutions and would guide health workers in administration of therapy to depressed students.