Neural Networks for Artificial Immune Systems: LVQ for Detectors Construction

This paper presents a non-standard approach for solving computer viruses detection problem based on the artificial immune system (AIS) method. The AIS is the biologically-inspired technique which have powerful information processing capabilities that makes it attractive for applying in computer security systems. Computer security systems based on AIS principles allow detect unknown malicious code. In this work we are describing model build on the AIS approach in which detectors represent the learning vector quantization (LVQ) neural networks. Basic principles of the biological immune system and comparative analysis of unknown computer viruses detection for different antivirus software and our model are presented.