FMR: Fast randomized algorithms for covariance matrix computations
COULAUD, Olivier
High-End Parallel Algorithms for Challenging Numerical Simulations [HiePACS]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
High-End Parallel Algorithms for Challenging Numerical Simulations [HiePACS]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
DARVE, Eric
Department of Mechanical Engineering [Stanford]
Institute for Computational and Mathematical Engineering [Stanford] [ICME]
See more >
Department of Mechanical Engineering [Stanford]
Institute for Computational and Mathematical Engineering [Stanford] [ICME]
COULAUD, Olivier
High-End Parallel Algorithms for Challenging Numerical Simulations [HiePACS]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
High-End Parallel Algorithms for Challenging Numerical Simulations [HiePACS]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
DARVE, Eric
Department of Mechanical Engineering [Stanford]
Institute for Computational and Mathematical Engineering [Stanford] [ICME]
Department of Mechanical Engineering [Stanford]
Institute for Computational and Mathematical Engineering [Stanford] [ICME]
FRANC, Alain
Biodiversité, Gènes & Communautés [BioGeCo]
from patterns to models in computational biodiversity and biotechnology [PLEIADE]
< Reduce
Biodiversité, Gènes & Communautés [BioGeCo]
from patterns to models in computational biodiversity and biotechnology [PLEIADE]
Language
en
Autre communication scientifique (congrès sans actes - poster - séminaire...)
This item was published in
Platform for Advanced Scientific Computing (PASC), 2016-06-08, Lausanne. 2016-06-08
English Abstract
We present an open-source library implementing fast algorithms for covari-ance matrices computations, e.g., randomized low-rank approximations (LRA) and fast multipole matrix multiplication (FMM). The library can be used ...Read more >
We present an open-source library implementing fast algorithms for covari-ance matrices computations, e.g., randomized low-rank approximations (LRA) and fast multipole matrix multiplication (FMM). The library can be used to approximate square roots of low-rank covariance matrices in O(N 2) operations in SVD form using randomized LRA, instead of the standard O(N 3) cost. Low-rank covariance matrices given as kernels, e.g., Gaussian decay, evaluated on 3D grids can be decomposed in O(N) operations using the FMM. The performance of the library is illustrated on two examples: • Generation of Gaussian Random Fields (GRF) on large spatial grids • MultiDimensional Scaling (MDS) for the classification of species.Read less <
Origin
Hal imported