Fast incremental expectation maximization for finite-sum optimization: nonasymptotic convergence.
Gersende FortP. GachEric MoulinesPublished in: Stat. Comput. (2021)
Keyphrases
- expectation maximization
- em algorithm
- evolutionary multiobjective
- optimization problems
- convergence rate
- optimization algorithm
- quasi newton
- global convergence
- line search
- mixture model
- combinatorial optimization
- constrained optimization
- faster convergence rate
- objective function
- gaussian mixture
- optimization process
- incremental learning
- global optimization
- optimization method
- generative model
- maximum likelihood
- probabilistic model
- evolutionary algorithm
- optimal solution
- gaussian mixture model
- parameter estimation
- weighted sum
- hidden markov models
- neural network