Defining and sampling conditioned diffusion processes. A member of the suite of packages from JuliaDiffusionBayes.
Key Features • Installation • How To Use • Related • License
Implementation of the Backward Filtering–Forward Guiding (BFFG) algorithm. It is a generic computational framework for working with Guided Proposals introduced by M Schauer, F van der Meulen and H van Zanten [arXiv].
The main object introduced by this package is a struct GuidProp and it allows for sampling of guided proposals, computing log-likelihood functions of the sampled trajectories and embedding the samplers in smoothing or inference algorithms.
The package is not yet registered. To install it, type in:
] add https://github.com/JuliaDiffusionBayes/GuidedProposals.jlSee the documentation.
GuidedProposals.jl belongs to a suite of packages in JuliaDiffusionBayes, whose aim is to facilitate Bayesian inference for diffusion processes. Some other packages in this suite are as follows:
- DiffusionDefinition.jl: define diffusion processes and sample from their laws
- ObservationSchemes.jl: a systematic way of encoding discrete-time observations for stochastic processes
- ExtensibleMCMC.jl: a modular implementation of the Markov chain Monte Carlo (MCMC) algorithms
- DiffusionMCMCTools.jl: utility methods that facilitate easier coding solutions for smoothing and inference algorithms for diffusions
- DiffusionMCMC.jl: Markov chain Monte Carlo (MCMC) algorithms for doing inference for diffusion processes
MIT