Towards Adaptive Self-Normalized Importance Samplers

Nicola Branchini, Víctor Elvira · 2025

Read Towards Adaptive Self-Normalized Importance Samplers by Nicola Branchini, Víctor Elvira (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-00372.txt.

Description

Nicola Branchini and Víctor Elvira improve a workhorse estimation method used across signal processing and machine learning. Self-normalized importance sampling guesses an average by drawing samples from a stand-in distribution and reweighting them. Everything hinges on picking a good stand-in, which is usually out of reach.

Earlier adaptive schemes tuned the stand-in for a different, unnormalized version of the problem. This work is the first to aim adaptation squarely at the best stand-in for the self-normalized case, using a chain of random samples to approach it step by step. The authors uncover a close kinship with a rival adaptive method for ratios, which suggests merging tricks from both camps. Numerical examples show the gains.

Catalog record

CREATORNicola Branchini, Víctor Elvira
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE20 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-00372

Browse the full library · About the Tower of Babel