S. Paradiso, M. Bonici, M. Chen, W.J. Percival, G. D'Amico, H. Zhang and G. McGee
{"title":"Reducing nuisance prior sensitivity via non-linear reparameterization, with application to EFT analyses of large-scale structure","authors":"S. Paradiso, M. Bonici, M. Chen, W.J. Percival, G. D'Amico, H. Zhang and G. McGee","doi":"10.1088/1475-7516/2025/07/005","DOIUrl":null,"url":null,"abstract":"Many physical models contain nuisance parameters that quantify unknown properties of an experiment that are not of primary relevance. Typically, these cannot be measured except by fitting the models to the data from the experiment, requiring simultaneous measurement of interesting parameters that are our target of inference and nuisance terms that are not directly of interest. A recent example of this is fitting Effective Field Theory (EFT) models to large-scale structure (LSS) data to make cosmological inferences. These models have a large number of nuisance parameters that are typically correlated with cosmological parameters in the posterior, leading to strong dependence on the nuisance parameter priors. We introduce a reparametrization method that leverages Generalized Additive Models (GAMs) to decorrelate nuisance parameters from the parameters of interest in the likelihood, even in the presence of non-linear relationships. This reparametrization forms a natural basis within which to define priors that are independent between nuisance and target parameters: the separation means that the marginal posterior for cosmological parameters does not depend on simple priors placed on nuisance terms. In application to EFT models using LSS data, we demonstrate that the proposed approach leads to robust cosmological inference.","PeriodicalId":15445,"journal":{"name":"Journal of Cosmology and Astroparticle Physics","volume":"47 1","pages":""},"PeriodicalIF":5.9000,"publicationDate":"2025-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Cosmology and Astroparticle Physics","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1088/1475-7516/2025/07/005","RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ASTRONOMY & ASTROPHYSICS","Score":null,"Total":0}
引用次数: 0
Abstract
Many physical models contain nuisance parameters that quantify unknown properties of an experiment that are not of primary relevance. Typically, these cannot be measured except by fitting the models to the data from the experiment, requiring simultaneous measurement of interesting parameters that are our target of inference and nuisance terms that are not directly of interest. A recent example of this is fitting Effective Field Theory (EFT) models to large-scale structure (LSS) data to make cosmological inferences. These models have a large number of nuisance parameters that are typically correlated with cosmological parameters in the posterior, leading to strong dependence on the nuisance parameter priors. We introduce a reparametrization method that leverages Generalized Additive Models (GAMs) to decorrelate nuisance parameters from the parameters of interest in the likelihood, even in the presence of non-linear relationships. This reparametrization forms a natural basis within which to define priors that are independent between nuisance and target parameters: the separation means that the marginal posterior for cosmological parameters does not depend on simple priors placed on nuisance terms. In application to EFT models using LSS data, we demonstrate that the proposed approach leads to robust cosmological inference.
期刊介绍:
Journal of Cosmology and Astroparticle Physics (JCAP) encompasses theoretical, observational and experimental areas as well as computation and simulation. The journal covers the latest developments in the theory of all fundamental interactions and their cosmological implications (e.g. M-theory and cosmology, brane cosmology). JCAP''s coverage also includes topics such as formation, dynamics and clustering of galaxies, pre-galactic star formation, x-ray astronomy, radio astronomy, gravitational lensing, active galactic nuclei, intergalactic and interstellar matter.