Faster Least Squares Approximation

Abstract

Least squares approximation is a technique to find an approximate solution to a system of linear equations that has no exact solution. In a typical setting, one lets $n$ be the number of constraints and $d$ be the number of variables, with $n\gg d$. Then, existing exact methods find a solution vector in $O(n{d}^2)$ time. We present two randomized algorithms that provide very accurate relative-error approximations to the optimal value and the solution vector of a least squares approximation problem more rapidly than existing exact algorithms. Both of our algorithms preprocess the data with the Randomized Hadamard Transform. One then uniformly randomly samples constraints and solves the smaller problem on those constraints, and the other performs a sparse random projection and solves the smaller problem on those projected coordinates. In both cases, solving the smaller problem provides relative-error approximations, and, if $n$ is sufficiently larger than $d$, the approximate solution can be computed in $O(nd\log d)$ time.

25 pages; minor changes from previous version; this version will appear in Numerische Mathematik

Cited by in corpus (8)

• Sketching as a Tool for Numerical Linear Algebra
• Effective Resistances, Statistical Leverage, and Applications to Linear Equation Solving
• A Sparse Johnson--Lindenstrauss Transform
• A fast algorithm for computing minimal-norm solutions to underdetermined systems of linear equations
• Sharper Bounds for Regularized Data Fitting
• A fast randomized algorithm for orthogonal projection
• Improved Low-rank Matrix Decompositions via the Subsampled Randomized Hadamard Transform
• Random Projections for the Nonnegative Least-Squares Problem