Scaled Conjugate Gradients for Maximum Likelihood: An Empirical Comparison with the EM Algorithm.
Kristian KerstingNiels LandwehrPublished in: Probabilistic Graphical Models (2002)
Keyphrases
- em algorithm
- maximum likelihood
- expectation maximization
- mixture model
- parameter estimation
- maximum likelihood estimation
- likelihood function
- maximum a posteriori
- gaussian mixture model
- log likelihood
- gaussian mixture
- hyperparameters
- log likelihood function
- hidden variables
- generative model
- likelihood maximization
- incomplete data
- expectation maximisation
- parameter learning
- finite mixture model
- finite mixture models
- penalized likelihood
- pairwise
- density estimation
- probability density function