Prediction of Depression using Machine Learning Techniques: A Review of Existing Literature

Depression is a serious mental disorder that negatively affects the mental health of a person, such that the person thinks, behaves, and feels in a negative way that can lead to physical and emotional problems. It can be a life-threatening problem as a person in such situation abandons hope of good and see negative aspects of situations. As per World Health Organization report of 2019, approximately 3 million people all over the world were suffering from depression between the ages of 17–25 and 40–70. Machine learning (ML) models can be trained on a training dataset so it can predict whether a person is going to develop depression or not. In the past, many researchers have worked for the prediction of depression by using ML Techniques. The purpose of this paper is to review the performance of ML techniques for early prediction of depression in older people. Different classifiers have been used, including Bayes Net, logistic regression, multilayer perceptron, sequential minimal optimization, decision table, and random forest. Our findings reveal that a maximum accuracy of 89% and maximum precision of 0.95 can be achieved from a large dataset.