Title :
An adaptive detection of anomalies in user´s behavior
Author :
Sokolov, Artern M.
Author_Institution :
Int. Res. & Training Center of Informational Technol. & Syst., Kiev, Ukraine
Abstract :
We propose an adaptive approach to modeling user´s behavior in computer anomaly detection systems. As a base model, Markov chains with variable memory length are used. An adaptive version of the algorithm constructing a model that takes into account changes in user´s behavior is introduced. Experimental testing of the proposed approach is also provided.
Keywords :
Markov processes; security of data; Markov chains; adaptive detection; base model; computer anomaly detection systems; user behavior; Adaptive algorithm; Authentication; Computer security; Cryptography; Frequency conversion; Hidden Markov models; Learning automata; Protection; Resists; System testing;
Conference_Titel :
Neural Networks, 2003. Proceedings of the International Joint Conference on
Print_ISBN :
0-7803-7898-9
DOI :
10.1109/IJCNN.2003.1223947
