CONSISTENCY OF MARKOV CHAIN QUASI-MONTE CARLO ON CONTINUOUS STATE SPACES By
نویسندگان
چکیده
The random numbers driving Markov chain Monte Carlo (MCMC) simulation are usually modeled as independent U(0, 1) random variables. Tribble [24] reports substantial improvements when those random numbers are replaced by carefully balanced inputs from completely uniformly distributed sequences. The previous theoretical justification for using anything other than IID U(0, 1) points shows consistency for estimated means, but only applies for discrete stationary distributions. We extend those results to some MCMC algorithms for continuous stationary distributions. The main motivation is the search for quasi-Monte Carlo versions of MCMC. As a side benefit, the results also establish consistency for the usual method of using pseudo-random numbers in place of random ones.
منابع مشابه
Consistency of Markov chain quasi-Monte Carlo on continuous state spaces
The random numbers driving Markov chain Monte Carlo (MCMC) simulation are usually modeled as independent U(0, 1) random variables. Tribble [24] reports substantial improvements when those random numbers are replaced by carefully balanced inputs from completely uniformly distributed sequences. The previous theoretical justification for using anything other than IID U(0, 1) points shows consisten...
متن کاملStream 1 Consistency of Markov chain quasi - Monte Carlo on continuous state spaces
Multivariate problems occur in many applications. They are defined on spaces of functions of d variables and d is often huge. Many of such multivariate problems suffer from the curse of dimensionality. This means that the minimal number of functions values needed to compute by algorithms with error at most ε is an exponential function of d. In this talk we discuss how the curse of dimensionalit...
متن کاملQuasi-Monte Carlo methods for Markov chains with continuous multi-dimensional state space
We describe a quasi-Monte Carlo method for the simulation of discrete time Markov chains with continuous multi-dimensional state space. The method simulates copies of the chain in parallel. At each step the copies are reordered according to their successive coordinates. We prove the convergence of the method when the number of copies increases. We illustrate the method with numerical examples w...
متن کاملLocal Quasi-Monte Carlo Exploration
In physically-based image synthesis, the path space of light transport paths is usually explored by stochastic sampling. The two main families of algorithms are Monte Carlo/quasi-Monte Carlo sampling and Markov chain Monte Carlo. While the former is known for good uniform discovery of important regions, the latter facilitates efficient exploration of local effects. We introduce a hybrid samplin...
متن کاملThe semantic structure of quasi-Borel spaces
Quasi-Borel spaces are a new mathematical structure that supports higher-order probability theory, first-order iteration, and modular semantic validation of Bayesian inference algorithms with continuous distributions. Like a measurable space, a quasi-Borel space is a set with extra structure suitable for defining probability and measure distributions. But unlikemeasurable spaces, quasi-Borel sp...
متن کامل