Décomposition de problèmes à variables multiples au moyen de la programmation génétique

Les détecteurs d'intrusion à base comportementale nécessitent de la formation afin de caractériser correctement l'exploitation d'une combinaison service¬–protocole. Cela implique que l'algorithme d'apprentissage soit applicable à d'importants ensembles de données pour fournir des solutions simples. Ce projet abordera ces deux exigences dans le contexte de la programmation génétique au moyen d'un modèle combiné hôte-parasite à […]

Read More
Multi-class Problem Decomposition Using Genetic Programming

Behavioural detectors for intrusion detection require training in order to correctly characterize the operation of a service – protocol combination. Implicit in this is the assumption that the learning algorithm will scale to large datasets and provide simple solutions. This work will address both requirements under a Genetic Programming context through the use of a […]

Read More
Correlating Intrusion Scenarios with an Unsupervised Learning Model

The increasing sophistication of distributed attacks on networked infrastructure has resulted in a requirement for tools capable of abstracting and alerting network managers of network status across multiple data sources. The basic objective of this project is to provide a framework for correlating information from multiple network sources into a cohesive picture of system status. […]

Read More