Directional data classification using a hierarchical model of von Mises distribution

The von Mises distribution1 VM -pdf is a continuous probability distribution on the circle used in directional statistics. A mixture model of von Mises distribution, which is broad enough to cover symmetry as well as asymmetry, unimodality as well as multimodality of circular data. In this paper we use a model comprised of a hierarchical von Mises mixture distribution mode HmvM- pdf where we consider each class is itself the result of a mixture of subclasses. The parameters of our model are estimated using the expectation maximization algorithm EM modified. The HmvM- pdf model achieves higher accuracy than the mvM model and offer a rich modeling. The suitability of the distributions is judged from the coefficient of determination R2. © 2017 Association for Computing Machinery.