Optimality in Noisy Importance Sampling
Machine Learning
2022-01-24 v1 Machine Learning
Computation
Abstract
In this work, we analyze the noisy importance sampling (IS), i.e., IS working with noisy evaluations of the target density. We present the general framework and derive optimal proposal densities for noisy IS estimators. The optimal proposals incorporate the information of the variance of the noisy realizations, proposing points in regions where the noise power is higher. We also compare the use of the optimal proposals with previous optimality approaches considered in a noisy IS framework.
Cite
@article{arxiv.2201.02432,
title = {Optimality in Noisy Importance Sampling},
author = {Fernando Llorente and Luca Martino and Jesse Read and David Delgado-Gómez},
journal= {arXiv preprint arXiv:2201.02432},
year = {2022}
}