Accelerating astronomical and cosmological inference with preconditioned Monte Carlo

Karamanis, Minas (avtor), Beutler, Florian (avtor), Peacock, John A. (avtor), Nabergoj, David (avtor), Seljak, Uroš (avtor)

We introduce preconditioned Monte Carlo (PMC), a novel Monte Carlo method for Bayesian inference that facilitates efficient sampling of probability distributions with non-trivial geometry. PMC utilizes a Normalizing Flow (NF) in order to decorrelate the parameters of the distribution and then proceeds by sampling from the preconditioned target distribution using an adaptive Sequential Monte Carlo (SMC) scheme. The results produced by PMC include samples from the posterior distribution and an estimate of the model evidence that can be used for parameter inference and model comparison, respectively. The aforementioned framework has been thoroughly tested in a variety of challenging target distributions achieving state-of-the-art sampling performance. In the cases of primordial feature analysis and gravitational wave inference, PMC is approximately 50 and 25 times faster, respectively, than nested sampling (NS). We found that in higher dimensional applications, the acceleration is even greater. Finally, PMC is directly parallelisable, manifesting linear scaling up to thousands of CPUs.

• Jezik: Angleški jezik
• Ključne besede: data analysis, statistics, large-scale structure of the Universe
• Tipologija: 1.01 - Izvirni znanstveni članek
• Organizacija: FRI - Fakulteta za računalništvo in informatiko
• Status publikacije: Objavljeno
• Različica publikacije: Objavljena publikacija
• Datum objave: 01.10.2022
• Leto izida: 2022
• Št. strani: Str. 1644-1653
• Številčenje: Vol. 516, iss. 2
• PID: 20.500.12556/RUL-171138
• UDK: 524.8:536.911
• ISSN pri članku: 0035-8711
• DOI: 10.1093/mnras/stac2272
• COBISS.SI-ID: 243782403
• Datum objave v RUL: 08.08.2025

Gradivo je del revije

• Naslov: Monthly notices of the royal astronomical society
• Skrajšan naslov: Mon. Not. R. Astron. Soc.
• Založnik: Blackwell Scientific Publications
• ISSN: 0035-8711
• COBISS.SI-ID: 25980416

Licence

• Licenca: CC BY 4.0, Creative Commons Priznanje avtorstva 4.0 Mednarodna
• Povezava: http://creativecommons.org/licenses/by/4.0/deed.sl
• Opis: To je standardna licenca Creative Commons, ki daje uporabnikom največ možnosti za nadaljnjo uporabo dela, pri čemer morajo navesti avtorja.

Sekundarni jezik

• Jezik: Slovenski jezik
• Ključne besede: analiza podatkov, statistika, struktura Vesolja

Projekti

• Financer: Drugi - Drug financer ali več financerjev
• Številka projekta: 853291

• Financer: Drugi - Drug financer ali več financerjev
• Številka projekta: DE-AC02-05CH11231