An anomaly-detection scheme is presented here in which the monitored process is modeled as a Markov chain. It generates an alarm depending on the likelihoods that the monitored process exhibits intrusive behavior or normal behavior. The scheme maximizes the true alarm rate such the fraction of alarms which are false does not exceed a given limit. The false alarm rate can be computed based on a model of normal behavior, which can be extracted from a sample of normal behavior. The true alarm rate, on the other hand, requires a model of intrusive behavior, even though samples of intrusive behavior are not widely available. Instead, a novel game-theoretical approach is used to model intrusive behavior also based on the sample of normal behavior. The uncertainties in both models are approximated by normal distribution, which simplifies the analysis and provides more insight.
By: Mehdi Nassehi
Published in: RZ3011 in 1998
LIMITED DISTRIBUTION NOTICE:
This Research Report is available. This report has been submitted for publication outside of IBM and will probably be copyrighted if accepted for publication. It has been issued as a Research Report for early dissemination of its contents. In view of the transfer of copyright to the outside publisher, its distribution outside of IBM prior to publication should be limited to peer communications and specific requests. After outside publication, requests should be filled only by reprints or legally obtained copies of the article (e.g., payment of royalties). I have read and understand this notice and am a member of the scientific community outside or inside of IBM seeking a single copy only.
Questions about this service can be mailed to reports@us.ibm.com .