Minimum Message Length Inference and Mixture Modelling of Inverse Gaussian Distributions.
Daniel F. SchmidtEnes MakalicPublished in: Australasian Conference on Artificial Intelligence (2012)
Keyphrases
- gaussian distribution
- minimum message length
- expectation maximization
- finite mixtures
- em algorithm
- gaussian mixture model
- maximum likelihood
- mixture model
- noise model
- unsupervised learning
- covariance matrices
- gaussian model
- multivariate gaussian
- mixture of gaussians
- kl divergence
- probabilistic model
- maximum likelihood estimation
- bayesian inference
- posterior distribution
- gaussian mixture
- data mining
- bayesian framework
- finite mixture model
- bayesian networks
- structured prediction
- parameter estimation
- active learning
- decision trees
- density estimation
- markov chain monte carlo
- object recognition
- dirichlet process
- multi variate
- objective function