On Semi-supervised Estimation of Discrete Distributions under f-divergences

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Erol, Hasan Sabri Melihcan, Zheng, Lizhong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911877389352960
author Erol, Hasan Sabri Melihcan
Zheng, Lizhong
author_facet Erol, Hasan Sabri Melihcan
Zheng, Lizhong
contents We study the problem of estimating the joint probability mass function (pmf) over two random variables. In particular, the estimation is based on the observation of $m$ samples containing both variables and $n$ samples missing one fixed variable. We adopt the minimax framework with $l^p_p$ loss functions. Recent work established that univariate minimax estimator combinations achieve minimax risk with the optimal first-order constant for $p \ge 2$ in the regime $m = o(n)$, questions remained for $p \le 2$ and various $f$-divergences. In our study, we affirm that these composite estimators are indeed minimax optimal for $l^p_p$ loss functions, specifically for the range $1 \le p \le 2$, including the critical $l_1$ loss. Additionally, we ascertain their optimality for a suite of $f$-divergences, such as KL, $χ^2$, Squared Hellinger, and Le Cam divergences.
format Preprint
id arxiv_https___arxiv_org_abs_2405_09523
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle On Semi-supervised Estimation of Discrete Distributions under f-divergences
Erol, Hasan Sabri Melihcan
Zheng, Lizhong
Statistics Theory
Information Theory
We study the problem of estimating the joint probability mass function (pmf) over two random variables. In particular, the estimation is based on the observation of $m$ samples containing both variables and $n$ samples missing one fixed variable. We adopt the minimax framework with $l^p_p$ loss functions. Recent work established that univariate minimax estimator combinations achieve minimax risk with the optimal first-order constant for $p \ge 2$ in the regime $m = o(n)$, questions remained for $p \le 2$ and various $f$-divergences. In our study, we affirm that these composite estimators are indeed minimax optimal for $l^p_p$ loss functions, specifically for the range $1 \le p \le 2$, including the critical $l_1$ loss. Additionally, we ascertain their optimality for a suite of $f$-divergences, such as KL, $χ^2$, Squared Hellinger, and Le Cam divergences.
title On Semi-supervised Estimation of Discrete Distributions under f-divergences
topic Statistics Theory
Information Theory
url https://arxiv.org/abs/2405.09523