Predicting Depression Among the University Students in the Covid-19 Pandemic Situation Using Machine Learning Algorithms

Abstract

Depression is a very common word term. But the problem of depression is not normal. People think it is a disease but it is not a disease. This is a mental problem. People suffer from depression for various reasons. At present, the amount of depression among the people seems to have increased a lot. We have tried to credit depression among all the university students of Bangladesh during the Covid-19 pandemic period. When someone is depressed we can find out what they want to do or what they have done to get rid of depression. Doctors say depression is a mental disorder. The person who is suffering from this problem is basically worried about his own wrong doing or any failure in life or there is no one close to him anymore and he cannot accept it. Depression can also be caused by family strife or financial problems. Depressed people don't want to talk to anyone else. They always try to hide their depression. In our research, we have used some scales to understand how a depressed student thinks about himself or what he wants to do and how he can free himself from depression. The survey will give us a clear idea about these issues. In this research Logistic Regression, Random Forest, Multi-Layer Perceptron, Multinomial Naïve Bayes, Gradient Boosting, K-Nearest Neighbours, AdaBoost, Support Vector Machine have been used. Among these algorithms, Random Forest Algorithm hasishown the bestiresult. Random Forest has aniaccuracy of 85.57% which is approximately 86%.