SAPPHIRE is a new preconditioned variance-reduced stochastic method with provable linear convergence for composite convex problems, achieving large speedups on regularized ERM.
Convergence rate analysis of a stochas- tic trust-region method via supermartingales.INFORMS Journal on Optimization, 1(2):92–119, 2019
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
SAPPHIRE: Preconditioned Stochastic Variance Reduction for Faster Large-Scale Statistical Learning
SAPPHIRE is a new preconditioned variance-reduced stochastic method with provable linear convergence for composite convex problems, achieving large speedups on regularized ERM.