The effects of combined application of SOM, ANFIS and Subtractive Clustering in detecting intrusions in computer networks
Building a system for the detection and prevention of intrusions into computer networks is a major challenge. Huge amounts of network traffic that process these systems are characterized by diversity and the data are described by a number of attributes. In addition, input data are often changing in a relatively short period of time, creating a completely new traffic patterns. This significantly complicates the identification of potentially unwanted network traffic. The aim of this paper is to present and analyze the effects of combined application of Self Organizing Map (SOM), Adaptive Neuro Fuzzy Inference System (ANFIS), Subtractive Clustering (SC) and Voting Mechanism (VM) in building systems for intrusion detection in computer networks in order to maintain an acceptable level of efficiency of data processing and increased system adaptivity.