Intrusion Detection and Intrusion Prevention Using Machine Learning and Genetic Algorithms

The purpose of the system is to detect infiltration,detect and identify attacks, and diagnose securityfailures in a computer system or networks and notifysecurity managers. The obstacles and problems involvedin designing an effective intrusion detection system canbe a large amount of data on computer network traffic,low detection rates and the production of wrong alerts,which create highly pessimistic systems and ultimatelydisregard for professionals. The system will be warned.In this paper, using machine learning techniques, theselection of the feature based on the genetic algorithm toselect the most effective features and the decision treewas used to teach the model. For testing, 10% of thestandard KDD Cup 99 dataset was used and MATLABsoftware was used. The results indicate that theproposed method of the 22 attack classes would detect21 attack classes and reach an accurate 97% detectionrate.

کلیدواژه ها:

computer networks ، penetration ، machine learning and genetic algorithm