Are Gibbs-Type Priors the Most Natural Generalization of the Dirichlet Process?
Pierpaolo De BlasiStefano FavaroAntonio LijoiRamsés H. MenaIgor PrünsterMatteo RuggieroPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2015)
Keyphrases
- dirichlet process
- markov chain monte carlo
- gaussian processes
- multi task learning
- mixture model
- nonparametric bayesian
- bayesian model
- sampling algorithm
- posterior distribution
- bayesian framework
- maximum a posteriori
- latent topics
- variational inference
- monte carlo
- parameter estimation
- generative model
- maximum likelihood
- variational methods
- probabilistic model
- expectation maximization
- markov random field
- co occurrence
- multi class