Accelerated Sampling for the Indian Buffet Process

Abstract

We often seek to identify co-occurring hidden features in a set of observations. The Indian Buffet Process (IBP) provides a non-parametric prior on the features present in each observation, but current inference techniques for the IBP often scale poorly. The collapsed Gibbs sampler for the IBP has a running time cubic in the number of observations, and the uncollapsed Gibbs sampler, while linear, is often slow to mix. We present a new linear-time collapsed Gibbs sampler for conjugate likelihood models and demonstrate its efficacy on large real-world datasets.

Cite

Text

Doshi-Velez and Ghahramani. "Accelerated Sampling for the Indian Buffet Process." International Conference on Machine Learning, 2009. doi:10.1145/1553374.1553409

Markdown

[Doshi-Velez and Ghahramani. "Accelerated Sampling for the Indian Buffet Process." International Conference on Machine Learning, 2009.](https://mlanthology.org/icml/2009/doshivelez2009icml-accelerated/) doi:10.1145/1553374.1553409

BibTeX

@inproceedings{doshivelez2009icml-accelerated,
  title     = {{Accelerated Sampling for the Indian Buffet Process}},
  author    = {Doshi-Velez, Finale and Ghahramani, Zoubin},
  booktitle = {International Conference on Machine Learning},
  year      = {2009},
  pages     = {273-280},
  doi       = {10.1145/1553374.1553409},
  url       = {https://mlanthology.org/icml/2009/doshivelez2009icml-accelerated/}
}