An Intelligent Test Method of Distributed Network Inbreak Based on Cluster Analysis

Abstract

Due to the lack of large-scale processing of network data in the process of distributed network inbreak intelligent detection, the detection accuracy is low. Therefore, a distributed network inbreak intelligent test method based on cluster analysis is proposed. The distributed network data is preprocessed through data attribute feature transformation, data normalization, data standardization and data dimensionality reduction. According to the preprocessing results, the distributed network data collection is divided into multiple clusters using the fuzzy K-means clustering algorithm., compute the remove from each cluster centre to other data targets, use Euclidean remove to construct the goal function of partition quality, extract distributed network inbreak data, determine the cost function of the convolutional neural networks-gate recurrent unit network model, and use the stochastic gradient descent algorithm to The convolutional neural networks-gate recurrent unit network model is trained, and the extracted distributed network inbreak data is input as an initial sample into the trained convolutional neural networks-gate recurrent unit network model, the model is solved, and the distributed network inbreak intelligent detection results are output. The analysis of the experimental results shows that the proposed method has higher precision and better detection effect in the intelligent detection of distributed network inbreak.