Graphics processing units and high-dimensional optimization
From MaRDI portal
multidimensional scalingnonnegative matrix factorizationparallel computingblock relaxationEM and MM algorithmsPET scanning
Software, source code, etc. for problems pertaining to statistics (62-04) Parallel numerical computation (65Y05) Computer graphics; computational geometry (digital and algorithmic aspects) (68U05) Parallel algorithms in computer science (68W10) Numerical algorithms for specific classes of architectures (65Y10)
Abstract: This paper discusses the potential of graphics processing units (GPUs) in high-dimensional optimization problems. A single GPU card with hundreds of arithmetic cores can be inserted in a personal computer and dramatically accelerates many statistical algorithms. To exploit these devices fully, optimization algorithms should reduce to multiple parallel tasks, each accessing a limited amount of data. These criteria favor EM and MM algorithms that separate parameters and data. To a lesser extent block relaxation and coordinate descent and ascent also qualify. We demonstrate the utility of GPUs in nonnegative matrix factorization, PET image reconstruction, and multidimensional scaling. Speedups of 100 fold can easily be attained. Over the next decade, GPUs will fundamentally alter the landscape of computational statistics. It is time for more statisticians to get on-board.
Recommendations
- Megapixel topology optimization on a graphics processing unit
- Parallel global optimization on GPU
- GPU implementation of a multiobjective search algorithm
- Computing of high breakdown regression estimators without sorting on graphics processing units
- GPU accelerated greedy algorithms for compressed sensing
Cites work
- scientific article; zbMATH DE number 3567782 (Why is no real title available?)
- scientific article; zbMATH DE number 3579883 (Why is no real title available?)
- scientific article; zbMATH DE number 1086057 (Why is no real title available?)
- scientific article; zbMATH DE number 1086058 (Why is no real title available?)
- scientific article; zbMATH DE number 756188 (Why is no real title available?)
- scientific article; zbMATH DE number 788227 (Why is no real title available?)
- A Statistical Model for Positron Emission Tomography
- A quasi-Newton acceleration for high-dimensional optimization algorithms
- Algorithms and applications for approximate nonnegative matrix factorization
- Composition Markov chains of multinomial type
- Conjugate Gradient Acceleration of the EM Algorithm
- Horseshoes in multidimensional scaling and local kernel methods
- Learning the parts of objects by non-negative matrix factorization
- Maximum likelihood estimation via the ECM algorithm: A general framework
- Parallel multivariate slice sampling
- Simple and Globally Convergent Methods for Accelerating the Convergence of Any EM Algorithm
- Squared polynomial extrapolation methods with cycling: an application to the positron emission tomography problem
- The ECME algorithm: A simple extension of EM and ECM with faster monotone convergence
- The EM Algorithm and Extensions, 2E
- The MM alternative to EM
- The tunneling method for global optimization in multidimensional scaling
Cited in
(22)- High-performance statistical computing in the computing environments of the 2020s
- GPU implementation of a multiobjective search algorithm
- Massive parallelization of serial inference algorithms for a complex generalized linear model
- MM algorithms for geometric and signomial programming
- Massive Parallelization of Massive Sample-Size Survival Analysis
- Megapixel topology optimization on a graphics processing unit
- Variational algorithms for biclustering models
- Multilevel quantile function modeling with application to birth outcomes
- Computing of high breakdown regression estimators without sorting on graphics processing units
- Massive Parallelization Boosts Big Bayesian Multidimensional Scaling
- scientific article; zbMATH DE number 177435 (Why is no real title available?)
- scientific article; zbMATH DE number 6731238 (Why is no real title available?)
- An Asynchronous Distributed Expectation Maximization Algorithm for Massive Data: The DEM Algorithm
- Scalable Bayesian inference for self-excitatory stochastic processes applied to big American gunfire data
- High-performance processing of covariance matrices using GPU computations
- Graphics card as a cheap supercomputer.
- GPU accelerated MCMC for modeling terrorist activity
- Parallel statistical computing for statistical inference
- GPU parameter tuning for tall and skinny dense linear least squares problems
- HIGH PRECISION INTEGER ADDITION, SUBTRACTION AND MULTIPLICATION WITH A GRAPHICS PROCESSING UNIT
- Estimation for general birth-death processes
- Towards a parallel component in a GPU-CUDA environment: a case study with the L-BFGS Harwell routine
This page was built for publication: Graphics processing units and high-dimensional optimization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q906530)
