Poisson Random Fields for Dynamic Feature Models
This work addresses the need for Bayesian nonparametric models that can capture feature dependencies over time, particularly for applications like analyzing evolving text corpora, though it appears incremental as it builds on existing Indian buffet process frameworks.
The authors tackled the problem of modeling time-dependent data with latent features by introducing the Wright-Fisher Indian buffet process (WF-IBP), a probabilistic model that uses Poisson random fields to construct dependent beta processes for feature allocation over time. They applied this model to a nonparametric topic analysis of the full NIPS paper corpus from 1987 to 2015, demonstrating its utility in handling temporal dependencies in text data.
We present the Wright-Fisher Indian buffet process (WF-IBP), a probabilistic model for time-dependent data assumed to have been generated by an unknown number of latent features. This model is suitable as a prior in Bayesian nonparametric feature allocation models in which the features underlying the observed data exhibit a dependency structure over time. More specifically, we establish a new framework for generating dependent Indian buffet processes, where the Poisson random field model from population genetics is used as a way of constructing dependent beta processes. Inference in the model is complex, and we describe a sophisticated Markov Chain Monte Carlo algorithm for exact posterior simulation. We apply our construction to develop a nonparametric focused topic model for collections of time-stamped text documents and test it on the full corpus of NIPS papers published from 1987 to 2015.