Evolutionary Memetic Models for Malware Intrusion Detection: A Comparative Quest for Computational Solution and Convergence

Abstract

Data security is now a pertinent issue with advent of the Internet. Methods like cryptography, firewalls and gateways used to prevent attacks on data are becoming unsuccessful. Thus, the need for Intrusion Detection System to enhance security efforts. Varying machine learning models are implemented for rule-based IDS using DARPA dataset to train and generate rules for classification via supportconfidence framework and a common fitness function to judge quality of each rule. This will help detect network anomalies, new attack types via rules and allow their addition into knowledgebase. Study presents results of the various stochastic models used with an aim to improve data security and integrity for networked resources.