An Approach for Social Networking Sites to Detect Depression from Various Machine Learning Techniques

Every stage of life is essential for mental health from infancy through adolescence to adulthood. Depression is a typical and critical mental health problem. Everyone occasionally feels down, but depression is different from common grief or loss. It affects attitudes, emotions, and actions. Additionally, it has a significant impact on how people handle tension, interact with others, and make wise choices. Physical and emotional health are both components of one's total health. For example, melancholy increases the chance of bodily health problems over the long run, particularly chronic conditions like diabetes, heart disease, and stroke. Current health facility practice is to collect the necessary information for depression diagnosis through several tests, and then give suitable therapy depending on the diagnosis. Big data analytics and machine learning approaches are efficient for collecting the huge dataset of the depression and predict the outcomes respectively. In this research, three distinct machine learning algorithms is used predict depression. The confusion matrix was found to be unbalanced after applying the various approaches. As a result, the f1 score was included, which aided in choosing the most suitable accuracy model among the three applied algorithms as Multilayer Perceptron.