Large scale inference in the Infinite Relational Model: Gibbs sampling is not enough.
Kristoffer Jon AlbersAndreas Leon Aagaard MothMorten MørupMikkel N. SchmidtPublished in: MLSP (2013)
Keyphrases
- gibbs sampling
- markov chain
- topic models
- gibbs sampler
- approximate inference
- belief networks
- exact inference
- em algorithm
- parameter estimation
- expectation maximization
- latent dirichlet allocation
- collapsed gibbs sampling
- markov chain monte carlo
- graphical models
- probabilistic inference
- belief propagation
- bayesian networks
- sampling methods
- structured prediction