Presents a self-normalized subsampling procedure for asymptotically valid confidence regions from SGD iterates under both finite and infinite variance assumptions.
Eliminating sharp minima from sgd with truncated heavy-tailed noise.arXiv preprint arXiv:2102.04297, 2021
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Statistical Inference for Stochastic Gradient Descent Beyond Finite Variance
Presents a self-normalized subsampling procedure for asymptotically valid confidence regions from SGD iterates under both finite and infinite variance assumptions.