This study aims to analyze the elements that affect the computer network security, then design the corresponding protection system, and detail its implementation mode. Based on artificial intelligence technology, it is committed to realizing the collection, analysis and processing of network data, so as to boost the effective application of Internet security and provide theoretical support for network engineering. The rapid development of Internet information has greatly improved the life and production of residents, but also caused many computer network security risks. Therefore, it is imperative to analyze and formulate the computer network security management strategy. Based on this study, by exploring the causes of computer network security, we propose the widely used machine learning security management technology. First, the design principles, overall architecture and network security structure of machine learning are discussed in depth, and then the support vector machine (SVM) algorithm, BP neural network algorithm and Web technology are introduced in detail. Combined with the advantages of machine learning intelligence and accuracy, the level of predictive analysis technology of computer network security management is expounded. Finally, the implementation and future goals of network security management technology are described, and the scientific development prospect of computer network security management based on machine learning is discussed.