I
Ian Porteous
Researcher at University of California, Irvine
Publications - 11
Citations - 1024
Ian Porteous is an academic researcher from University of California, Irvine. The author has contributed to research in topics: Gibbs sampling & Mixture model. The author has an hindex of 7, co-authored 11 publications receiving 967 citations.
Papers
More filters
Proceedings ArticleDOI
Fast collapsed gibbs sampling for latent dirichlet allocation
TL;DR: A novel collapsed Gibbs sampling method for the widely used latent Dirichlet allocation (LDA) model, which can be as much as 8 times faster than the standard collapsed Gibbs sampler for LDA and results in significant speedups on real world text corpora.
Proceedings Article
Bayesian matrix factorization with side information and dirichlet process mixtures
TL;DR: A Bayesian matrix factorization model that performs regression against side information known about the data in addition to the observations is introduced and applied to the Netflix Prize problem of predicting movie ratings given an incomplete user-movie ratings matrix.
Proceedings ArticleDOI
Unsupervised learning of visual taxonomies
TL;DR: The experiments show that a disorganized collection of images will be organized into an intuitive taxonomy and it is found that the taxonomy allows good image categorization and, in this respect, is superior to the popular LDA model.
Proceedings Article
Multi-HDP: a non parametric Bayesian model for tensor factorization
TL;DR: This work introduces a novel generative Bayesian probabilistic model for unsupervised matrix and tensor factorization and describes an efficient collapsed Gibbs sampler for inference.
Proceedings Article
Gibbs sampling for (Coupled) infinite mixture models in the stick breaking representation
TL;DR: In this paper, Gibbs samplers for infinite complexity mixture models in the stick breaking representation are explored to improve mixing over cluster labels and to bring clusters into correspondence, and an application to modeling of storm trajectories is used to illustrate these ideas.