Title :
Dynamic Evolution Systems and Applications in Intrusion Detection Systems
Author :
Xu, Xian-Ming ; Zhan, Justin
Author_Institution :
Carnegie Mellon CyLab, Kobe
Abstract :
In this paper, we present a dynamic evolution system and build up a model to trace the transition of the system state. This new model differs from the previous methods, such as Bayesian network, artificial neural network, in two aspects: it can adapt the changes of the environment automatically, and it does not need a special training phase to build up a model. Theoretical analysis shows that it is applicable and practical, and furthermore, experimental results show that it has good performance especially in dynamic environment.
Keywords :
security of data; dynamic environment; dynamic evolution system; intrusion detection system; system state transition; Artificial neural networks; Bayesian methods; Data security; Detectors; Information security; Intrusion detection; Open systems; Performance analysis; Testing; Training data; dynamic; evolution; intrusion;
Conference_Titel :
Information Security and Assurance, 2008. ISA 2008. International Conference on
Conference_Location :
Busan
Print_ISBN :
978-0-7695-3126-7
DOI :
10.1109/ISA.2008.82
