Retracted Article: Collaborative correlation space big data clustering algorithm for abnormal flow monitoring

Abstract

The big data clustering process is a random nonlinear process with high uncertainty. Because traditional methods require prior knowledge to learn, they cannot adapt well to the real-time changes of big data, and cannot effectively achieve big data clustering. A good clustering structure can reduce redundancy, optimize network resource configuration, and reduce node overhead and balance the network. The collaborative correlation space is a powerful tool that will simulate the model to form a spatial analysis and process simulation. Therefore, in order to improve the fast processing and recognition ability of big data, a collaborative correlation spatial big data oriented to clustering network is proposed. Simulation experiments show that using this algorithm for big data clustering can effectively improve the data clustering efficiency, reduce energy consumption, has better anti-interference and adaptability, and has higher clustering accuracy. In the flow anomalydetectionexperiment,resultsshowthatthemethodproposedinthispaperhashighertrafficanomaly identificationaccuracythank-meansanddecisiontreealgorithm,andtherecallrateandROCareaarethelargest.