Psychological Crisis Prediction of Students Based on Network Behavior by Big Data Mining

Signs of psychological crisis can be found in time by analyzing the network behavior data of college students, thus providing a basis for early warning and intervention. However, existing methods may not only have shortcomings in handling dynamic data and updating models, but also rely too much on network behavior data and overlook other factors possibly affecting the psychological crisis of college students. In order to overcome these shortcomings, this paper aimed to study the psychological crisis prediction of college students based on big data mining of network behavior. Network behavior interactive prediction was defined to determine the objective function of the constructed model. Interactive prediction model framework was presented and the working principle of the model was explained. Finally, various early warning indexes, which needed to be comprehensively considered in the psychological crisis early warning model of college students, were given, and the combination of principal component analysis (PCA) and support vector machine (SVM) was applied to the construction of the early warning model, thus improving its prediction effects, generalization ability and interpretability, and reducing the overfitting risk and the difficulty of processing high-dimensional data. The experimental results verified that the constructed model was effective.