Hidden Markov Models (HMMs) and Security Applications

Abstract

The Hidden Markov models (HMMs) are statistical models used in various communities and applications. Such applications include speech recognition, mental task classification, biological analysis, and anomaly detection. In hidden Markov models, there are two states: one is a hidden state and the other is an observation state. The purpose of this survey paper is to further the understanding of hidden Markov models, as well as the solutions to the three central problems: evaluation problem, decoding problem and learning problem. In addition, applying HMMs in real world applications such as security and engineering will improve the classification and accuracy for the whole field.