Conditionally structured variational Gaussian approximation with importance weights.
Linda S. L. TanAishwarya BhaskaranDavid J. NottPublished in: Stat. Comput. (2020)
Keyphrases
- relative importance
- gaussian convolution
- linear combination
- series expansion
- linear computational complexity
- weight assignment
- weight function
- independent and identically distributed
- weighted sum
- taylor series expansion
- closed form
- statistically independent
- image segmentation
- polynomial approximation
- approximation algorithms
- weighting scheme
- variational methods
- free energy
- error bounds
- structured data
- maximum likelihood
- zero crossing
- gaussian model
- approximation methods
- gaussian mixture
- variational approximation
- gaussian mixture model
- lower bound
- image sequences
- real world