DF-GMVI approximates Bayesian posteriors with Gaussian mixtures using derivative-free quadrature that preserves covariance positivity and affine invariance.
Amari , Natural gradient works efficiently in learning , Neural computation, 10 (1998), pp
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Stable Derivative Free Gaussian Mixture Variational Inference for Bayesian Inverse Problems
DF-GMVI approximates Bayesian posteriors with Gaussian mixtures using derivative-free quadrature that preserves covariance positivity and affine invariance.