I2 Tieto- ja viestintätekniset sovellukset
AdaptiveMCMC.jl (2020)
Vihola, M. (2020). AdaptiveMCMC.jl. GitHub. https://github.com/mvihola/AdaptiveMCMC.jl
JYU-tekijät tai -toimittajat
Julkaisun tiedot
Julkaisun kaikki tekijät tai toimittajat: Vihola, Matti
Julkaisuvuosi: 2020
Ilmestymispäivä: 11.10.2021
Kustantaja: GitHub
Julkaisun kieli: englanti
Pysyvä verkko-osoite: https://github.com/mvihola/AdaptiveMCMC.jl
Julkaisun avoin saatavuus: Avoimesti saatavilla
Julkaisukanavan avoin saatavuus: Kokonaan avoin julkaisukanava
Tiivistelmä
This package provides implementations of some general-purpose random-walk based adaptive MCMC algorithms, including the following:
Adaptive Metropolis, proposal covariance adaptation, (Haario, Saksman and Tamminen, 2001, and Andrieu and Moulines, 2006)
Adaptive scaling Metropolis, acceptance rate adaptation for scale (e.g. as in Andrieu and Thoms, 2008, and Atchadé and Fort, 2010)
Robust Adaptive Metropolis, acceptance rate adaptation for shape (Vihola, 2012)
Adaptive Parallel Tempering, acceptance rate adaptation for temperature levels (Miasojedow, Moulines and Vihola, 2013)
The aim of the package is to provide a simple and modular general-purpose implementation, which may be easily used to sample from a log-target density, but also used in a variety of custom settings.
Adaptive Metropolis, proposal covariance adaptation, (Haario, Saksman and Tamminen, 2001, and Andrieu and Moulines, 2006)
Adaptive scaling Metropolis, acceptance rate adaptation for scale (e.g. as in Andrieu and Thoms, 2008, and Atchadé and Fort, 2010)
Robust Adaptive Metropolis, acceptance rate adaptation for shape (Vihola, 2012)
Adaptive Parallel Tempering, acceptance rate adaptation for temperature levels (Miasojedow, Moulines and Vihola, 2013)
The aim of the package is to provide a simple and modular general-purpose implementation, which may be easily used to sample from a log-target density, but also used in a variety of custom settings.
YSO-asiasanat: otanta; stokastiset prosessit; Markovin ketjut; Monte Carlo -menetelmät; algoritmit
Liittyvät organisaatiot
OKM-raportointi: Kyllä
