Application And Analysis Of Convolutional Neural Network Algorithm For Security Defense Mechanism

With the rapid development of the Internet and data science,the application and processing demand of electronic information data in cloud computing and Internet is increasing day by day.In addition,it is necessary to prevent the possible risks of data,such as unauthorized access,computer virus and network attack,so as to reduce the hidden security risks of computer network system.These mixed hazards greatly affect people's use experience and personal privacy data security,and with the continuous growth of applications,network attacks emerge in an endless stream,seriously threatening the security of the entire Internet environment in the cyberspace.By summarizing the experience and lessons of previous major network attacks,the attack risks can be extracted and used for reference by the defense.Combined with the solution of data security measures,the information security of the system platform can be properly guaranteed.How to effectively establish security defense mechanism against data risks has become an important topic widely concerned in the field of information security nowadays.Convolutional neural network is a kind of feedforward neural network,and its characteristics can be found from shallow to deep after data is processed by convolutional layer and pooling layer.In this paper,the computer access service flow is oriented.By crawling and filtering the access data source to the access service,the convolutional neural network is used to discover the hazard characteristics layer by layer and identify the attack category,so as to make the corresponding defense processing.Focusing on the problems of unauthorized access of Internet data,computer virus and network attack,and aiming at the security risk of information application in the environment of computer platform,an intrusion detection mechanism design for information application was proposed,and the security threat source data was tested and analyzed by combining the convolutional neural network algorithm.The CNN modeling analysis is compared with the traditional artificial neural network,naive bayes and recursive neural network algorithm,and it is found that the accuracy of this algorithm has certain advantages over other algorithms.Finally,GPU calculation is introduced to speed up the processing of convolutional neural network algorithm combined with a security example,so as to effectively improve the operation efficiency of the security defense mechanism.