Stochastic gradient ascent with a diagonal empirical-Fisher per-coordinate scale learns Dirichlet process mixture posteriors on par with batch coordinate ascent on image datasets up to 108K images.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Fisher Information based Stochastic Gradient Ascent for Online Learning of Dirichlet Process Mixture and Theory
Stochastic gradient ascent with a diagonal empirical-Fisher per-coordinate scale learns Dirichlet process mixture posteriors on par with batch coordinate ascent on image datasets up to 108K images.