Introduces a consistent estimator for the sub-Gaussian variance proxy via constrained empirical maximization of L(λ), with rates O_p(n^{-1/2+ε}) or O_p(n^{-1/2}) under assumptions on L, minimax optimality in a subclass, and behavior for non-sub-Gaussian cases.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Estimation of the sub-Gaussian parameter
Introduces a consistent estimator for the sub-Gaussian variance proxy via constrained empirical maximization of L(λ), with rates O_p(n^{-1/2+ε}) or O_p(n^{-1/2}) under assumptions on L, minimax optimality in a subclass, and behavior for non-sub-Gaussian cases.