CoLFI: cosmological likelihood-free inference with neural density estimators

Simple item page
dc.contributor.authorBeesham, Aroonkumar
dc.contributor.authorWang, Guo-jian
dc.contributor.authorCheng, Cheng
dc.contributor.authorMa, Yin-zhe
dc.contributor.authorXia, Jun-qing
dc.contributor.authorAbebe, Amare
dc.coverage.conferenceissn
dc.date.accessioned2025-12-03T09:06:01Z
dc.date.available2025-12-03T09:06:01Z
dc.date.issued2023
dc.departmentNameMathematical Sciences
dc.description.abstractIn previous works, we proposed to estimate cosmological parameters with an artificial neural network (ANN) and a mixture density network (MDN). In this work, we propose an improved method called a mixture neural network (MNN) to achieve parameter estimation by combining ANN and MDN, which can overcome shortcomings of the ANN and MDN methods. Besides, we propose sampling parameters in a hyperellipsoid for the generation of the training set, which makes the parameter estimation more efficient. A high-fidelity posterior distribution can be obtained using ( ) 102 forward simulation samples. In addition, we develop a code named CoLFI for parameter estimation, which incorporates the advantages of MNN, ANN, and MDN, and is suitable for any parameter estimation of complicated models in a wide range of scientific fields. CoLFI provides a more efficient way for parameter estimation, especially for cases where the likelihood function is intractable or cosmological models are complex and resource-consuming. It can learn the conditional probability density p(θ|d) using samples generated by models, and the posterior distribution p(θ|d0) can be obtained for a given observational data d0. We tested the MNN using power spectra of the cosmic microwave background and Type Ia supernovae and obtained almost the same result as the Markov Chain Monte Carlo method. The numerical difference only exists at the level of ( ) s  - 10 2 . The method can be extended to higher-dimensional data.
dc.facultyFaculty of Science, Agriculture and Engineering
dc.format.preprintNo
dc.identifier.citationWang, G.J., Cheng, C., Ma, Y.Z., Xia, J.Q., Abebe, A. and Beesham, A., 2023. CoLFI: cosmological likelihood-free inference with neural density estimators. The Astrophysical Journal Supplement Series, 268(1), pp.1-20.
dc.identifier.issn1538-4365 (online)
dc.identifier.issn0067-0049 (print)
dc.identifier.otherhttps://doi.org/10.3847/1538-4365/ace113
dc.identifier.urihttp://hdl.handle.net/10530/58475
dc.inproceedingsissn
dc.issuenumber268 / 7
dc.keynoteissn
dc.language.isoen
dc.pages1 - 20
dc.peerreviewedYes
dc.publisherAmerican Astronomical Society
dc.subjectArtificial neural network
dc.subjectMixture density network
dc.subjectCosmology
dc.titleCoLFI: cosmological likelihood-free inference with neural density estimators
dc.title.journalThe Astrophysical Journal Supplement Series
dc.typeJournal Article
dspace.entity.typePublication
relation.isAuthorOfPublicationcac21718-4edb-4a03-a532-568ac9799cc1
relation.isAuthorOfPublication.latestForDiscoverycac21718-4edb-4a03-a532-568ac9799cc1
relation.isOrgUnitOfPublication71c06a0e-e111-424c-8caf-20609eacb82c
relation.isOrgUnitOfPublication.latestForDiscovery71c06a0e-e111-424c-8caf-20609eacb82c
Files:
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Beesham_Aroonkumar_2023.pdf
Size:
3.03 MB
Format:
Adobe Portable Document Format

Beesham_Aroonkumar_2023.pdf (3.03 MB)

Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:

license.txt (1.71 KB)

Download
Collections: Articles