Sparse Stochastic Inference for Latent Dirichlet allocation
arXiv:1206.6425
Abstract
We present a hybrid algorithm for Bayesian topic models that combines the efficiency of sparse Gibbs sampling with the scalability of online stochastic inference. We used our algorithm to analyze a corpus of 1.2 million books (33 billion words) with thousands of topics. Our approach reduces the bias of variational inference and generalizes to many Bayesian hidden-variable models.
Appears in Proceedings of the 29th International Conference on Machine Learning (ICML 2012)
Cited by in corpus (7)
- Structured Stochastic Variational Inference
- Stochastic Variational Inference
- Scalable and Robust Construction of Topical Hierarchies
- An Empirical Study of Stochastic Variational Algorithms for the Beta Bernoulli Process
- Managing sparsity, time, and quality of inference in topic models
- Online Inference for Relation Extraction with a Reduced Feature Set
- Efficient Correlated Topic Modeling with Topic Embedding