Network Intrusion Detection Based On Clustering And Random Forest Technology Research

The Internet has developed rapidly in recent years,has important influence on the production and living of people,greatly promoted the development of the society.But at the same time,related to network attack and destruction of produce continuously,the mode of attack and link emerge in endlessly,seriously affected the normal operation of the Internet users,and this issue has caused more and more people's attention.Because all sorts of attack means of constant development,the original traditional network protection system has been difficult to achieve safe operation of the Internet,the network defense tends to simplification of the static,one-sided.Therefore,intrusion detection technology has been rapid development trend in this network,will become important content in the field of technology research and development in the future.This approach greatly changed the traditional research content and means,is not a passive safety monitoring,but active monitoring of the invasion,become an important supplementary content network security field.Rapid progress in the development of the Internet,in the process of the new network and for invasive monitoring must be improved monitoring method and validity.And the original detection method is hard to adapt to the new period of monitoring requirements,leading to greater likelihood of detection error,for network security loopholes and hidden trouble.Combined with the problems existing in the current detection,the author thought that should be innovative means of intrusion detection,clustering analysis and random forests are existing reasonable monitoring method,it can be to combine,comprehensive utilization of the advantages of both improved monitoring of validity.First of all,in this paper,the intrusion detection research status at home and abroad,and in this paper,the research background,method and content.Secondly,carries on the related theory research,explore intrusion detection research status and stage of explain the concepts involved.Again,the emphasis on the characteristics of clustering analysis algorithm and the using method,with the method of data,the traditional approach to innovation,improve data availability.Clustering analysis the advantages and strengths,the objective to identify the shortcoming of this algorithm,explore the actual model of clustering analysis algorithm to generate ideas about the rationality of clustering analysis;Then,combined with random forests in the role of intrusion detection technology,studies the characteristics of random forest detection method,the corresponding model was constructed and its level of intrusion detection and validity.Finally,this paper research on the combination of the two algorithms put forward the feasibility proposal,then combined with the advantages of both algorithms in specific monitoring,disadvantages,enhance the efficiency of monitoring,monitoring in the field of Internet security for the future was discussed.