Research on Network Security Situation Awareness System Based on Machine Learning

摘要: With the popularity of computers, Internet has entered production and all aspects social life, but attendant problem network security become focus widespread concern. Network situation awareness to effectively respond issues provide a viable solution: for complex environments malicious attacks, comprehensive analysis attacks against various parts system, from macro point view be assess predict future based on this information. For predictive accuracy prediction system improved significantly, method machine learning have high degree feasible, in real applications certain research practical value. Introduction Security Situational Awareness is current status trends entire information variety factors operating conditions devices, behavior user constituted. situational can acquire, understand display environment elements. Through series technical means time space, are fully aware access associated elements as possible pluralistic, establishment behaviors modeling simulation pre-side then integrate analyze vast amounts with network-related data. scientific effective assessment use relevant technologies make reasonable predictions about over security, management personnel advance remind equipment, peer node hosts data resources adjustments, upgrades backup, address risk harm losses may result down an acceptable range . Extract carried out basis that only collection sophisticated index ensure correctness results assessment. Therefore, design process model or must pay attention select metric system. source diverse, different methods, devices collect formats, letter mainly contains configuration, status, traffic, other