Stochastic approximation algorithms for partition function estimation of Gibbs random fields.
Gerasimos PotamianosJohn K. GoutsiasPublished in: IEEE Trans. Inf. Theory (1997)
Keyphrases
- approximation algorithms
- partition function
- gibbs random fields
- importance sampling
- special case
- np hard
- maximum likelihood estimation
- monte carlo
- worst case
- parameter estimation
- upper bound
- closed form
- markov random field
- minimum cost
- lower bound
- random fields
- upper and lower bounds
- approximate inference
- probability density function
- structure learning
- maximum likelihood
- structured prediction
- constant factor approximation
- higher order
- combinatorial auctions
- belief propagation
- probability distribution