Bayesian Estimation of the von-Mises Fisher Mixture Model with Variational Inference.
Jalil TaghiaZhanyu MaArne LeijonPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2014)
Keyphrases
- variational inference
- von mises fisher
- mixture model
- posterior distribution
- expectation maximization
- hyperparameters
- em algorithm
- model selection
- probabilistic model
- gaussian mixture model
- generative model
- maximum likelihood
- latent variables
- density estimation
- parameter estimation
- bayesian framework
- posterior probability
- probability distribution
- hidden variables
- unsupervised learning
- maximum a posteriori
- language model
- exponential family
- cross validation
- probability density function
- unit sphere
- gaussian distribution
- bayesian inference
- topic models
- information retrieval
- text classification
- multiscale
- training data