A deterministic Kaczmarz algorithm for solving linear systems

05/17/2021
by   Changpeng Shao, et al.
0

We propose a deterministic Kaczmarz method for solving linear systems A=$̱ withAnonsingular. Instead of using orthogonal projections, we use reflections in the original Kaczmarz iterative method. This generates a series of points on ann-spherecentered at the solution_*=A^-1$̱. We show that these points are nicely distributed on . Taking the average of several points will lead to an effective approximation to the solution. We will show how to choose these points efficiently. The numerical tests show that in practice this deterministic scheme converges much faster than we expected and can beat the (block) randomized Kaczmarz methods.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
06/01/2020

Randomized Kaczmarz for Tensor Linear Systems

Solving linear systems of equations is a fundamental problem in mathemat...
research
11/18/2022

A greedy randomized average block projection method for linear feasibility problems

The randomized projection (RP) method is a simple iterative scheme for s...
research
08/01/2019

A Randomized Algorithm for Preconditioner Selection

The task of choosing a preconditioner M to use when solving a linear sys...
research
04/11/2019

A Kaczmarz Algorithm for Solving Tree Based Distributed Systems of Equations

The Kaczmarz algorithm is an iterative method for solving systems of lin...
research
09/05/2022

A negative dependence framework to assess different forms of scrambling

We use the framework of dependence to assess the benefits of scrambling ...
research
02/19/2020

On the Planar Two-Center Problem and Circular Hulls

Given a set S of n points in the Euclidean plane, the two-center problem...
research
05/29/2017

Coreset Construction via Randomized Matrix Multiplication

Coresets are small sets of points that approximate the properties of a l...

Please sign up or login with your details

Forgot password? Click here to reset