Early Depression Prediction among Nigerian University Students Using Adaptive Neuro-Fuzzy Inference System (ANFIS)

Depression is a mental disorder characterized by a sad mood, irritability, anger, agitations, loss of interest or pleasure, reduced energy, feelings of guilt, low self-esteem, troubled sleep, appetite loss, and poor attentiveness. The effects of late diagnosis of depression in Nigerian students have posed threats to the academic performance of the students, economic growth, and security threats. To address this challenge, an ANFIS model for early detection of depression among Nigerian Students is proposed. This aids in the reduction and possible elimination of prevalent cases of depression-related dangers among students in tertiary institutions. ANFIS is utilized because of its transparency and ability to classify and identify hidden symptoms of depression, and its tendency for reduced memorization errors for users. The database was developed to hold user data, symptoms, and prescriptions and linked to the ANFIS framework to enable the diagnosis of early-phase depression. Data was collected from the University of Uyo primary health care center, and the University of Uyo Teaching Hospital (UUTH). The ANFIS model implementation was implemented in MATLAB while the application forming the input interface was implemented with JAVA. The dataset for training was passed through ANFIS for 10 epochs and upon completion the system had a training error of 6.0138e-0.5 and an average testing error of 4.6648 on the test data, these results indicate that the system possessed 95% classification accuracy in the detection of early depression in Nigerian students.