The convergent Indian buffet process
This addresses a theoretical limitation in Bayesian nonparametric priors for researchers in machine learning and statistics, though it appears incremental as it modifies an existing process.
The authors tackled the problem of unbounded feature growth in latent feature models by proposing the convergent Indian buffet process (CIBP), which ensures the expected number of features converges to a bounded value as objects increase, unlike the standard process where it grows indefinitely.
We propose a new Bayesian nonparametric prior for latent feature models, which we call the convergent Indian buffet process (CIBP). We show that under the CIBP, the number of latent features is distributed as a Poisson distribution with the mean monotonically increasing but converging to a certain value as the number of objects goes to infinity. That is, the expected number of features is bounded above even when the number of objects goes to infinity, unlike the standard Indian buffet process under which the expected number of features increases with the number of objects. We provide two alternative representations of the CIBP based on a hierarchical distribution and a completely random measure, respectively, which are of independent interest. The proposed CIBP is assessed on a high-dimensional sparse factor model.