Security Analysis of Power System User Behavior in Big Data Based on Comparative of Neural Network Methods

In the context of exponential growth of data information, a large amount of data needs to be effectively processed in the power system intranet to analyze and obtain valuable information. Among them, user behavior is especially important for the security of the entire power system. With the development of large data technology and the wide application of machine learning technology, useful information can be extracted and processed from a large amount of unprocessed data. At the same time, in order to reasonably expand the data set of the relevant samples, traditional mathematical modeling can preprocess the data and integrate the classified output into the sample data set. Finally, this paper selects SSH data as the training set and Local data as the test set, and compares the accuracy of the four machine learning methods in predicting user behavior categories. The ultimate goal is to predict the level of risk of user behavior through the characteristics of user behavior.