Enhanced intrusion network system using Fuzzy –K-Mediod clustering method

Abstract

Intrusion detection scheme (IDS) is software applications that are used for monitoring of the network to recognize the malicious activity in the system. With the advent in technology and internet, the incident of the intruder activity on the system has been increased. Hence, the protection and security of the system become an essential approach because attackers utilize different kinds of attack methods to hack the useful information. However, various intrusion detection methods and algorithm have been developed to detect various types of the attacks. Some issues in existing research are excesses of information with different volumetric data. It became difficult to detect intrusion in large amount of data in a computer network.. Hence, the proposed research on different machine learning approach is used for network intrusion detection. In addition, clustering method implemented to segment the features using k mean clustering and k-medoids clustering algorithm. Moreover, implement an enhanced Fuzzy k medoids clustering approach for recognition of intrusion and faults on the network. Fuzzy k-mediod clustering helps in the evaluation of the maximum degree matrix. Experimental analysis is done by evaluating and comparing the parameters using Precision, Recall, Accuracy,FAR and FRR.