The Infinite Gamma-Poisson Feature Model
Michalis K. Titsias · 2007
We present a probability distribution over non-negative integer valued matrices with possibly an infinite number of columns. We also derive a stochastic process that reproduces this distribution over equivalence classes. This model can play the role of the prior in nonparametric Bayesian learning scenarios where multiple latent features are associated with the observed data and each feature can have multiple appearances or occurrences within each data point. Such data arise nat-urally when learning visual object recognition systems from unlabelled images. Together with the nonparametric prior we consider a likelihood model that ex-plains the visual appearance and location of local image patches. Inference with this model is carried out using a Markov chain Monte Carlo algorithm. 1