Analysis Method of Abnormal Traffic of Teaching Network in Higher Vocational Massive Open Online Course Based on Deep Convolutional Neural Network

Abstract

AbstractThe running security and stability of massive open online course's teaching network directly affect the implementation process of massive open online course's teaching tasks. In order to provide valuable reference data for the management and maintenance of the network, this paper puts forward an analysis method of abnormal traffic of higher vocational massive open online course's teaching network based on deep convolution neural network. According to the structure of the teaching network of massive open online course in higher vocational colleges, a network model is built. Under this model, the flow data is collected, and the preprocessing of the initial flow data is completed through data cleaning, standardized conversion and clustering. The deep convolution neural network is established, and the characteristics of network traffic data are extracted through back propagation iteration. After discrete detection and feature matching, the abnormal traffic in the teaching network of higher vocational massive open online course is detected, and the visual analysis results are obtained. Compared with the traditional network abnormal traffic analysis method, it is found that the detection error and missed detection rate of the optimized design method are reduced by 3.75MB and 1.02% respectively, the accuracy of traffic abnormal type analysis is increased by 1%, and the analysis speed is obviously improved.KeywordsDeep Convolution Neural NetworkTeaching in Massive Open Online CourseTeaching NetworkAbnormal Flow Analysis