athomas_jtsamplers

Chordal graph samplers

Package

Alun Thomas

Version

76ad20e

Language

Java

Docs

Paper

Giudici and Green[1], Green and Thomas[2]

Graph type

DG

Docker

bpimages/athomas_jtsamplers:9eb505f

Module folder

athomas_jtsamplers

Description

Abstract [2]: Full Bayesian computational inference for model determination in undirected graphical models is currently restricted to decomposable graphs or other special cases, except for small-scale problems, say up to 15 variables. In this paper we develop new, more efficient methodology for such inference, by making two contributions to the computational geometry of decomposable graphs. The first of these provides sufficient conditions under which it is possible to completely connect two disconnected complete subsets of vertices, or perform the reverse procedure, yet maintain decomposability of the graph. The second is a new Markov chain Monte Carlo sampler for arbitrary positive distributions on decomposable graphs, taking a junction tree representing the graph as its state variable.

Some fields described

  • edge_penalty Set the edge penalty in the prior.

  • num_samples The number of MCMC iterations.

  • sampler Set the sampler to the one indexed by: 0 = Giudicci & Green (1999) sampler. 1 = Green & Thomas (2013) single edge junction tree sampler. 2 = Green & Thomas (2013) multiple edge junction tree sampler.

  • size_maxclique Set the maximum clique size.

Example JSON

[
  {
    "id": "gg99",
    "burnin_frac": 0.5,
    "mcmc_estimator": "map",
    "timeout": null,
    "threshold": 0.5,
    "mcmc_seed": 1,
    "num_samples": 10000,
    "sampler": 0,
    "edge_penalty": 1.0,
    "size_maxclique": 10000
  }
]