Langevin and hessian with fisher approximation stochastic sampling for parameter estimation of structured covariance.
Cornelia Paula VacarJean-François GiovannelliYannick BerthoumieuPublished in: ICASSP (2011)
Keyphrases
- parameter estimation
- stochastic sampling
- importance sampling
- markov chain monte carlo
- approximate inference
- maximum likelihood
- posterior distribution
- markov random field
- least squares
- model selection
- computationally inexpensive
- simulated annealing
- em algorithm
- monte carlo
- covariance matrices
- parameter estimation algorithm
- parameter values
- expectation maximization
- markov chain
- genetic algorithm
- solution space
- partial solutions
- covariance matrix
- particle filtering
- objective function
- three dimensional