Federated Expectation Maximization with heterogeneity mitigation and variance reduction.
Aymeric DieuleveutGersende FortEric MoulinesGeneviève RobinPublished in: CoRR (2021)
Keyphrases
- variance reduction
- expectation maximization
- em algorithm
- gradient estimation
- monte carlo
- sample size
- maximum likelihood
- probabilistic model
- mixture model
- bias variance decomposition
- random numbers
- parameter estimation
- importance sampling
- generative model
- maximum a posteriori
- naive bayes classifier
- quasi monte carlo
- probability density function
- gaussian distribution
- image segmentation
- bayesian framework
- distributed systems
- k means
- bayesian networks
- posterior distribution
- confidence intervals
- graphical models
- text mining
- trade off