CausalNet4IDS is a research project for reasoning cyber intrusions using a Bayesian networks approach. We created a novel detection framework leveraing Copula Bayesian networks model, which is regarded an efficient way for both parameter estimation and strucutre learning. The resultant directed acyclic graph encodes causal relationships among variables, that is, relevant information describing why and how an attack happens in modern cyber intrusion detection systems.
- Matlab 2013a or higher.
- BNT toolbox
- Visualize we use graphviz
- see file matlab/causal_analysis.m
- data are stored in folder datasets/
Department of Computer Science, Technical University of Munich Boltzmannstr.3 , 85748 Garching (near Munich), Germany
Homepage: https://www.sec.in.tum.de
Copyright 2015-10 Huang Xiao and George Webster