Geom-SPIDER-EM: Faster Variance Reduced Stochastic Expectation Maximization for Nonconvex Finite-Sum Optimization.
Gersende FortEric MoulinesHoi-To WaiPublished in: CoRR (2020)
Keyphrases
- expectation maximization
- em algorithm
- global optimization
- optimization problems
- probabilistic model
- mixture model
- maximum likelihood
- gaussian mixture model
- generative model
- parameter estimation
- image segmentation
- unsupervised learning
- nonlinear programming
- probability density function
- bayesian framework
- objective function
- stochastic search
- stochastic optimization
- gaussian mixture
- stochastic programming
- optimization algorithm
- maximum a posteriori
- discrete random variables
- k means
- maximum likelihood estimation
- expectation maximisation
- gaussian distribution
- convex optimization
- evolutionary algorithm
- globally convergent
- monte carlo
- constrained optimization
- likelihood function
- particle swarm optimization
- correlation coefficient
- web crawling
- min sum
- log likelihood function
- covariance matrix