A variational component splitting approach for finite generalized Dirichlet mixture models

In this paper, a component splitting and local model selection method is proposed to address the mission of learning and selecting generalized Dirichlet (GD) mixture model with feature selection in an incremental variational way. Under the proposed principled variational framework, we simultaneously estimate, in a closed-form, all the involved parameters and determine the complexity (i.e. both model and features selection) of the GD mixture. The effectiveness of the proposed approach is evaluated using synthetic data as well as real a challenging application involving image categorization.