Anomaly detection using Markov chain model

摘要

This paper provides a method of mathematical representation of the traffic flow of network states. The flow of states is represented as transitions to the Markov Chains. Anomalies are interpreted as graph transitions with low probabilities.