Distributed and Adaptive Darting Monte Carlo through Regenerations

Darting Monte Carlo (DMC) is a MCMC procedure designed to effectively mix between multiple modes of a probability distribution. We propose an adaptive and distributed version of this method by using regenerations. This allows us to run multiple chains in parallel and adapt the shape of the jump regions as well as all other aspects of the Markov chain on the fly. We show that this significantly improves the performance of DMC because 1) a population of chains has a higher chance of finding the modes in the distribution, 2) jumping between modes becomes easier due to the adaptation of their shapes, 3) computation is much more efficient due to parallelization across multiple processors. While the curse of dimensionality is a challenge for both DMC and regeneration, we find that their combination ameliorates this issue slightly.

[1]  Bin Yu,et al.  Regeneration in Markov chain samplers , 1995 .

[2]  Cristian Sminchisescu,et al.  Generalized Darting Monte Carlo , 2007, AISTATS.

[3]  Nando de Freitas,et al.  Variational MCMC , 2001, UAI.

[4]  Max Welling,et al.  Accelerated Variational Dirichlet Process Mixtures , 2006, NIPS.

[5]  N. Metropolis,et al.  The Monte Carlo method. , 1949 .

[6]  Adrian F. M. Smith,et al.  Sampling-Based Approaches to Calculating Marginal Densities , 1990 .

[7]  J. Kadane,et al.  Identification of Regeneration Times in MCMC Simulation, With Application to Adaptive Schemes , 2005 .

[8]  Andrew Gelman,et al.  General methods for monitoring convergence of iterative simulations , 1998 .

[9]  John W. Fisher,et al.  Nonparametric belief propagation for self-localization of sensor networks , 2005, IEEE Journal on Selected Areas in Communications.

[10]  G. Roberts,et al.  Adaptive Markov Chain Monte Carlo through Regeneration , 1998 .

[11]  L. Tierney.  Markov Chains for Exploring Posterior Distributions , 1994 .

[12]  Kathryn B. Laskey,et al.  Population Markov Chain Monte Carlo , 2004, Machine Learning.

[13]  Cajo J. F. ter Braak,et al.  A Markov Chain Monte Carlo version of the genetic algorithm Differential Evolution: easy Bayesian computing for real parameter spaces , 2006, Stat. Comput..

[14]  Nando de Freitas,et al.  An Introduction to MCMC for Machine Learning , 2004, Machine Learning.

[15]  G. Warnes.  The Normal Kernel Coupler: An Adaptive Markov Chain Monte Carlo Method for Efficiently Sampling From Multi-Modal Distributions , 2001 .