Modelling and Mining of Networked Information Spaces

Project Title: Integration
Participants: Robert Curry, Dr. Malcolm Heywood
Project Description: Integration of attacker and detector is performed by way of co-evolution. The basic objective of which is to ensure that the detector and attacker develop at similar rates. Should one side dominate the other then evolution will come to a halt. The basic methodology investigated for achieving this will make use of the concept of exemplar dominance. That is to say, we are interested in identifying the subset of non-overlapping (attack) behaviours and matching this to the performance of detectors. Several approaches to investigating this problem will be modeled, including Nash equilibria, host-parasite models, and dynamic subset selection.