Properties of the stochastic approximation EM algorithm with mini-batch sampling.
Estelle KuhnCatherine MatiasTabea RebafkaPublished in: Stat. Comput. (2020)
Keyphrases
- em algorithm
- stochastic approximation
- expectation maximization
- monte carlo
- maximum likelihood
- mixture model
- parameter estimation
- generative model
- gaussian mixture model
- maximum likelihood estimation
- gaussian mixture
- likelihood function
- incomplete data
- expectation maximisation
- hyperparameters
- hidden variables
- penalized likelihood
- random sampling
- density estimation
- model based clustering
- sample size
- maximum a posteriori