Depression Detection of University Students' Using Machine Learning Approach

Abstract

In modern society, depression affects the majority of people. Many even commit suicide as a result of inadequate care. If depression is present in the patient in the early stages, it is simple to identify and cure. Since we don't know the depression level, we can't make the best choice at the appropriate time. A whole nation is built on its students. By teaching and improving the country, students represent it to the outside world. Depression is caused by a variety of factors, including challenges Bangladeshi adolescents have in their education. Determining the prevalence of depressive symptoms, their contributing variables, and strategies for reducing depression of university students are the objectives of our study. We analyzed the dataset with different samples from university students. We provide some question by a Google from, students are chosen the answer then set a range to find their depression level. About 1049 people's data were obtained from a Google form. In essence, the test was the data. In essence, the student has provided the data. We were able to determine the depression level through the analysis of that data. On this data, several algorithms have been applied. And have achieved the highest accuracy. With the help of this project, we can detect depression levels and administer the appropriate care or therapy. We're using some kind of algorithm to detect their depression level. They are five algorithms are chosen for this research. There are Decision Tree classification, Random Forest classifier, SVM, KNeighbors Classifier, and GaussianNB. Overall, the SVM algorithm prediction has the best performance. It gives 93% which is the best algorithm to prefer for this research.