Activity Recognition in Smart Homes Based on Second-Order Hidden Markov Model

Abstract

Hidden Markov Model is an important approach applied to activity recognition. In the first-order Hidden Markov Model, there is the hypothesis that the transition probability of state and the output probability of observation are only dependent on the current state of the model, which debases the precision of information extraction comparatively. In second-order Hidden Markov Model, the relevance between the current state and its previous two states is considered. Also, the relevance between the current observation and its previous state is considered. So second-order Hidden Markov Model has stronger performance of recognition of incorrect information. In my paper, second-order Hidden Markov Model is applied to activity recognition. The experiments show that our approach has higher precision than those approaches based on first-order Hidden Markov Model and based on Conditional Random Fields.