Parallelizing MCMC sampling via space partitioning

Abstract Efficient sampling of many-dimensional and multimodal density functions is a task great interest in many research fields. We describe an algorithm that allows parallelizing inherently serial Markov chain Monte Carlo (MCMC) by partitioning the space function parameters into multiple subspaces each them independently. The samples different are then reweighted their integral values stitched back together. This approach reducing wall-clock time parallel operation. It also improves target densities results less correlated samples. Finally, yields estimate function.