{"title":"A gradient adaptive population importance sampler","authors":"V. Elvira, Luca Martino, D. Luengo, J. Corander","doi":"10.1109/ICASSP.2015.7178737","DOIUrl":null,"url":null,"abstract":"Monte Carlo (MC) methods are widely used in signal processing and machine learning. A well-known class of MC methods is composed of importance sampling and its adaptive extensions (e.g., population Monte Carlo). In this paper, we introduce an adaptive importance sampler using a population of proposal densities. The novel algorithm dynamically optimizes the cloud of proposals, adapting them using information about the gradient and Hessian matrix of the target distribution. Moreover, a new kind of interaction in the adaptation of the proposal densities is introduced, establishing a trade-off between attaining a good performance in terms of mean square error and robustness to initialization.","PeriodicalId":117666,"journal":{"name":"2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2015-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"23","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICASSP.2015.7178737","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 23
Abstract
Monte Carlo (MC) methods are widely used in signal processing and machine learning. A well-known class of MC methods is composed of importance sampling and its adaptive extensions (e.g., population Monte Carlo). In this paper, we introduce an adaptive importance sampler using a population of proposal densities. The novel algorithm dynamically optimizes the cloud of proposals, adapting them using information about the gradient and Hessian matrix of the target distribution. Moreover, a new kind of interaction in the adaptation of the proposal densities is introduced, establishing a trade-off between attaining a good performance in terms of mean square error and robustness to initialization.