The Hidden Markov Model and its application to Human Activity Recognition

The process of predicting a qualitative response for an observation is referred to as classification. Supervised learning tools require labelled datasets to build classification models. However there are often instances when we have an unstructured dataset that doesn´t have the output sequence for the corresponding input sequence, i.e. the dataset available is unlabeled. In such cases we need to use Unsupervised Learning Tools to classify our data. Hidden Markov Model (HMM) is one such tool used to classify temporal data. Human activity recognition is one of the many areas where HMM can be used. In this paper we review the working of HMM as well as try to implement it for Human Activity Recognition.