An interruptible algorithm for perfect sampling via Markov chains

Fill, James Allen

doi:10.1214/aoap/1027961037

Cited by 133 publications

(128 citation statements)

References 26 publications

Supporting

Mentioning

127

Contrasting

Order By: Relevance

“…Various useful modifications of the basic MCMC have been suggested to address specific important applications. Notable among them are Green (1995), Besag and Clifford (1991), Fill (1998), Diaconis and Sturmfels (1998), Higdon (1998), and Kendall and Thönnes (1999), among many others. Various other specific references are given in the sections.…”

Section: Simulation and Markov Chain Monte Carlomentioning

confidence: 99%

Probability for Statistics and Machine Learning

Dasgupta¹

2011

Springer Texts in Statistics

View full text Add to dashboard Cite

Section: Simulation and Markov Chain Monte Carlomentioning

confidence: 99%

Probability for Statistics and Machine Learning

Dasgupta¹

2011

Springer Texts in Statistics

View full text Add to dashboard Cite

“…Perfect simulation of such models based on Gibbs sampling and extensions of Fill's algorithm (Fill, 1998) are studied in , Mùller and Schladitz (1999) and Wilson (2000). Below we consider perfect single-component slice sampling in the general case when all j j < I.…”

Section: Example 3: Ising and Other Automodelsmentioning

confidence: 99%

Perfect Slice Samplers

Mira

Møller

Roberts

2001

Journal of the Royal Statistical Society Series B: Statistical Methodology

View full text Add to dashboard Cite

Perfect sampling allows the exact simulation of random variables from the stationary measure of a Markov chain. By exploiting monotonicity properties of the slice sampler we show that a perfect version of the algorithm can be easily implemented, at least when the target distribution is bounded. Various extensions, including perfect product slice samplers, and examples of applications are discussed.

show abstract

“…Recently, Propp and Wilson [20] and this author [12] have devised algorithms to use Markov chain transitions to produce observations exactly from ir. Both algorithms work best when the transition matrix is monotone in a suitable sense we shall review.…”

Section: *-L N-r+l Kmentioning

confidence: 99%

“…An algorithm based on acceptance/rejection sampling and presented in Section 7.2 of [12], or rather a conceptually equivalent algorithm based on strong stationary duality and discussed in Section 9 of [12], is applied to M in Section 4. the Propp-Wilson algorithm is applied to M in Section 3, where it is shown that the method amounts to sampling with replacement but ignoring repetitions of record requests.…”

Section: *-L N-r+l Kmentioning

confidence: 99%

The Move-to-Front Rule: A Case Study for two Perfect Sampling Algorithms

Fill

1998

Prob. Eng. Inf. Sci.

Self Cite

View full text Add to dashboard Cite

The elementary problem of exhaustively sampling a finite population without replacement is used as a nonreversible test case for comparing two recently proposed MCMC algorithms for perfect sampling, one based on backward coupling and the other on strong stationary duality. The backward coupling algorithm runs faster in this case, but the duality-based algorithm is unbiased for user impatience. An interesting by-product of the analysis is a new and simple stochastic interpretation of a mixing-time result for the move-to-front rule.

show abstract

An interruptible algorithm for perfect sampling via Markov chains

Cited by 133 publications

References 26 publications

Probability for Statistics and Machine Learning

Probability for Statistics and Machine Learning

Perfect Slice Samplers

The Move-to-Front Rule: A Case Study for two Perfect Sampling Algorithms

Contact Info

Product

Resources

About