Cover image.

research

Anytime Monte Carlo

L.M. Murray, S. Singh, and A. Lee

Online

A Monte Carlo algorithm typically simulates some prescribed number of samples, taking some random real time to complete the computations necessary. This work considers the converse: to impose a real-time budget on the computation, so that the number of samples simulated is random. To complicate matters, the real time taken for each simulation may depend on the sample produced, so that the samples themselves are not independent of their number, and a length bias with respect to compute time is apparent. This is especially problematic when a Markov chain Monte Carlo (MCMC) algorithm is used and the final state of the Markov chain—rather than an average over all states—is required. The length bias does not diminish with the compute budget in this case. It occurs, for example, in sequential Monte Carlo (SMC) algorithms. We propose an anytime framework to address the concern, using a continuous-time Markov jump process to study the progress of the computation in real time. We show that the length bias can be eliminated for any MCMC algorithm by using a multiple chain construction. The utility of this construction is demonstrated on a large-scale SMC2^2 implementation, using four billion particles distributed across a cluster of 128 graphics processing units on the Amazon EC2 service. The anytime framework imposes a real-time budget on the MCMC move steps within SMC2^2, ensuring that all processors are simultaneously ready for the resampling step, demonstrably reducing wait times and providing substantial control over the total compute budget.

L.M. Murray, S. Singh, and A. Lee (2016). Anytime Monte Carlo.

L.M. Murray, S. Singh, and A. Lee (2016). <a href="https://indii.org/research/anytime-monte-carlo/">Anytime Monte Carlo</a>. 

@Article{,
  title = {Anytime {M}onte {C}arlo},
  author = {Lawrence M. Murray and Sumeetpal Singh and Pierre E. Jacob and Anthony Lee},
  year = {2016},
  url = {https://arxiv.org/abs/1612.03319}
}
blog Latest
GPU Programming in the Cloud
How to develop on remote cloud instances, and a roundup of cloud service providers.

Lawrence Murray

22 Nov 22

GPU Programming in the Cloud
blog Related
Matrix Gradients of Scalar Functions
Understanding the building blocks of reverse-mode automatic differentiation.

Lawrence Murray

7 Nov 22

photography Next
Winter Colours of Stockholm
research Previous
Comparative Analysis of Dengue and Zika Outbreaks Reveals Differences by Setting and Virus

S. Funk, A.J. Kucharski, A. Camacho, R.M. Eggo, L. Yakob, L.M. Murray, and W.J. Edmunds