Fast and Accurate Gaussian Kernel Ridge Regression Using Matrix Decompositions for Preconditioning

DOI10.1137/20M1343993MaRDI QIDQ5006445zbMATH OpenOpenAlexFDO

Authors Gil Shabat, Era Choshen, Dvir Ben Or, Nadav Carmel

Publication date 16 August 2021

Published in SIAM Journal on Matrix Analysis and Applications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1905.10587

zbMATH Keywords

CURinterpolative decompositionkernel ridge regressionpreconditionerrandomized matrix decompositions

Mathematics Subject Classification ID

Preconditioners for iterative methods (65F08) Iterative numerical methods for linear systems (65F10) Numerical linear algebra (65F99)

Recommendations

• Faster kernel ridge regression using sketching and preconditioning
• An adaptive factorized Nyström preconditioner for regularized kernel matrices
• Randomized Nyström Preconditioning
• Divide and conquer kernel ridge regression: a distributed algorithm with minimax optimal rates
• Efficient Gaussian process regression for large datasets

Cites work

• A fast algorithm for particle simulations
• A randomized algorithm for the decomposition of matrices
• A sparse Johnson-Lindenstrauss transform
• A theory of pseudoskeleton approximations
• Block basis factorization for scalable kernel evaluation
• Cover-based bounds on the numerical rank of Gaussian kernels
• CUR matrix decompositions for improved data analysis
• Efficient Algorithms for Computing a Strong Rank-Revealing QR Factorization
• Efficient algorithms for CUR and interpolative matrix decompositions
• Faster kernel ridge regression using sketching and preconditioning
• Finding structure with randomness: probabilistic algorithms for constructing approximate matrix decompositions
• scientific article; zbMATH DE number 3892457 (Why is no real title available?)
• Input sparsity time low-rank approximation via ridge leverage score sampling
• Matrix mathematics. Theory, facts, and formulas
• Multiscale data sampling and function extension
• Numerical Geometry of Non-Rigid Shapes
• On Rank-Revealing Factorisations
• On the Compression of Low Rank Matrices
• On the existence and computation of rank-revealing LU factorizations
• On the Nyström method for approximating a gram matrix for improved kernel-based learning
• Pseudo-skeleton approximations with better accuracy estimates
• Randomized Algorithms for Matrices and Data
• Randomized algorithms for the low-rank approximation of matrices
• Randomized LU decomposition
• Randomized LU decomposition using sparse projections
• Rank-Revealing QR Factorizations and the Singular Value Decomposition
• Relative-Error CUR Matrix Decompositions
• Strong rank revealing LU factorizations
• The Fast Gauss Transform

Cited in

(2)

• An adaptive factorized Nyström preconditioner for regularized kernel matrices
• Practical multi-fidelity machine learning: fusion of deterministic and Bayesian models

Describes a project that uses

Uses Software

• GPyTorch

This page was built for publication: Fast and Accurate Gaussian Kernel Ridge Regression Using Matrix Decompositions for Preconditioning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5006445)