Iterative numerical methods for sampling from high dimensional Gaussian distributions.
Erlend AuneJo EidsvikYvo PokernPublished in: Stat. Comput. (2013)
Keyphrases
- numerical methods
- gaussian distribution
- high dimensional
- multi variate
- parameter space
- maximum likelihood
- differential equations
- partial differential equations
- low dimensional
- dimensionality reduction
- covariance matrices
- gaussian mixture model
- finite element method
- sample size
- expectation maximization
- kl divergence
- data points
- high dimensional data
- multivariate gaussian
- level set method
- pairwise
- finite difference method
- runge kutta
- face recognition
- mixture model
- higher order
- probabilistic model
- dynamic programming
- feature space
- multiscale