Study of C4.5 algorithm based intrusion detection system
Zuo Yan-jie · Journal of Science of Teachers' College and University · 2007
Studied a decision tree to replace the existing structure of the chain structure intrusion detection system.It aimed at improving the storage and pattern matching to improve detection rate.Meanwhile,the use of C4.5. Decision tree algorithm for rule set each node dynamic selection of the most notable features parallel testing and achieve optimal performance. To minimize redundant and ineffective compared match, a cut on C4.5 decision tree algorithm,Information as to the criteria for judging rules. The experimental results show that this method can improve the forecast accuracy of the final model, and good overcome noise,effectively improved the speed and pattern matching rules access to the speed.