A Big Data Application To Predict Depression In The University Based On The Reading Habits

This study deeply mines the correlation between reading habits and depressive tendency of university students based on the data set from university library records and mental health questionnaires results. This paper compares different text categorization algorithms including kNN, SVM and naive Bayesian classifier on accuracy and time-consuming under different sample sizes. Ultimately, we construct a book classifier using naive Bayesian classification algorithm based on polynomial model, and the accuracy of the classifier reaches to 0.823. To build a psychological prediction model, this study uses linear regression algorithm and logistic regression algorithm respectively and the prediction accuracy of the models are compared under different error criteria, which turns out the logistic model behaves better.