PreprintarXiv licencearXiv
Instance-Specific Sharp Convergence Rate of the Add-λ Estimator via Feature Space Decomposition
Abstract
Using the Feature Space Decomposition framework, we establish non-asymptotic matching upper and lower bounds for the expected Kullback-Leibler estimation error of the Add-λ estimator of a discrete probability distribution, for any signal, any dimension, any sample size, and any λ>0.
Related
- Benign Overfitting Property of the Minimum ℓq -norm Interpolant Estimator in Regression Problem via Feature Space Decomposition
- Sharp convergence rate of the estimation error of LASSO via the Feature Space Decomposition method
- Impact of Feature Scaling on the Convergence Speed of Linear SVM Solvers : An Empirical Benchmark
- The skeleton-blocks decomposition of Bienaymé trees, and applications to their local convergence
- The geometric convergence of a parallel domain decomposition method on manifolds
BibTeX
@misc{shang2026instancespecific,
title = {{Instance-Specific Sharp Convergence Rate of the Add-$λ$ Estimator via Feature Space Decomposition}},
author = {Zong Shang},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01828},
eprint = {2611.01828},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01828},
}
RIS
TY - UNPB
TI - Instance-Specific Sharp Convergence Rate of the Add-λ Estimator via Feature Space Decomposition
AU - Zong Shang
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01828
UR - https://arxiv.org/abs/2611.01828
AB - Using the Feature Space Decomposition framework, we establish non-asymptotic matching upper and lower bounds for the expected Kullback-Leibler estimation error of the Add-λ estimator of a discrete probability distribution, for any signal, any dimension, any sample size, and any λ>0.
ER -