• Corpus ID: 88512788

Fully Parallel Particle Learning for GPGPUs and Other Parallel Devices

@article{McAlinn2012FullyPP,
  title={Fully Parallel Particle Learning for GPGPUs and Other Parallel Devices},
  author={Kenichiro McAlinn and Teruo Nakatsuma},
  journal={arXiv: Computation},
  year={2012}
}
We develop a novel parallel resampling algorithm for fully parallelized particle filters, which is designed with GPUs (graphics processing units) or similar parallel computing devices in mind. With our new algorithm, a full cycle of particle filtering (computing the value of the likelihood for each particle, constructing the cumulative distribution function (CDF) for resampling, resampling the particles with the CDF, and propagating new particles for the next cycle) can be executed in a… 

Figures and Tables from this paper

GPU-Powered Shotgun Stochastic Search for Dirichlet Process Mixtures of Gaussian Graphical Models
  • C. Mukherjee, Abel Rodríguez
  • Computer Science
    Journal of computational and graphical statistics : a joint publication of American Statistical Association, Institute of Mathematical Statistics, Interface Foundation of North America
  • 2016
TLDR
A novel stochastic search algorithm for finding the posterior mode of high-dimensional Dirichlet process mixtures of decomposable GGMs and investigates how to harness the massive thread-parallelization capabilities of graphical processing units to accelerate computation.
A hardware-accelerated particle filter for the geolocation of 1 demersal fishes 2 3
20 Geolocation is increasingly employed to reconstruct the movements of demersal fishes using data retrieved 21 from electronic archival tags. However, geolocation methods commonly suffer from
Pseudo-marginal algorithms with multiple CPUs
TLDR
This paper proposes to take advantage of multiple central processing units (CPUs), that are readily available on most standard desktop computers, and proposes a Bayesian algorithm that remains on the marginal space of the model parameter, with a target distribution that is still the correct posterior distribution.

References

SHOWING 1-10 OF 44 REFERENCES
A graphics processing unit implementation of the particle filter
TLDR
GPGPU techniques are used to make a parallel GPU implementation of state-of-the-art recursive Bayesian estimation using particle filters (PF), and the performance of the resulting GPU implementation is compared to one achieved with a traditional CPU implementation.
Particle Filtering: The Need for Speed
TLDR
GPGPU techniques are used to make a parallel recursive Bayesian estimation implementation using particle filters, which is the first complete GPU implementation of a PF published to this date, and is faster than the CPU filter when many particles are used, maintaining the same accuracy.
On the Utility of Graphics Cards to Perform Massively Parallel Simulation of Advanced Monte Carlo Methods
  • Anthony Lee, C. Yau, M. Giles, A. Doucet, C. Holmes
  • Computer Science
    Journal of computational and graphical statistics : a joint publication of American Statistical Association, Institute of Mathematical Statistics, Interface Foundation of North America
  • 2010
TLDR
It is suggested that GPUs have the potential to facilitate the growth of statistical modeling into complex data-rich domains through the availability of cheap and accessible many-core computation.
A Single Instruction Multiple Data Particle Filter
TLDR
This paper describes a Single Instruction Multiple Data (SIMD) implementation of a particle filter that uses N processors to process N particles and has a time complexity of O((log N)2) when performing resampling using N processors.
Massively Parallel Sequential Monte Carlo for Bayesian Inference
TLDR
This paper develops an algorithm that is well suited to parallel computing in general and for which convergence results have been established in the sequential Monte Carlo literature but that tends to require manual tuning in practical application, and introduces endogenous adaptations in the algorithm that obviate the need for tuning.
Resampling algorithms and architectures for distributed particle filters
TLDR
The proposed algorithms improve the scalability of the filter architectures affected by the resampling process and reduce communication through the interconnection network is reduced and made deterministic, which results in simpler network structure and increased sampling frequency.
Particle filter on GPUs for real-time tracking
TLDR
This work presents a new framework for particle filters based on discrete representations of probability densities, which takes advantage of knowledge about previous states of the system reducing the computational cost of an exhaustive search over the whole image.
Markov chain Monte Carlo, Sufficient Statistics, and Particle Filters
This article considers how to implement Markov chain Monte Carlo (MCMC) moves within a particle filter. Previous, similar, attempts have required the complete history (“trajectory”) of each particle
Sequential Monte Carlo Methods in Practice
TLDR
This book presents the first comprehensive treatment of Monte Carlo techniques, including convergence results and applications to tracking, guidance, automated target recognition, aircraft navigation, robot navigation, econometrics, financial modeling, neural networks, optimal control, optimal filtering, communications, reinforcement learning, signal enhancement, model averaging and selection.
Comparison of resampling schemes for particle filtering
  • R. Douc, O. Cappé, É. Moulines
  • Mathematics
    ISPA 2005. Proceedings of the 4th International Symposium on Image and Signal Processing and Analysis, 2005.
  • 2005
TLDR
It is first shown using simple arguments that the so-called residual and stratified methods do yield an improvement over the basic multinomial resampling approach, and a central limit theorem is established for the case where resamplings is performed using the residual approach.
...
...