Fast convergence to higher multiplicity zeros

11/25/2019
by   Sara Pollock, et al.
0

In this paper, the Newton-Anderson method, which results from applying an extrapolation technique known as Anderson acceleration to Newton's method, is shown both analytically and numerically to provide superlinear convergence to non-simple roots of scalar equations. The method requires neither a priori knowledge of the multiplicities of the roots, nor computation of any additional function evaluations or derivatives.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
07/25/2022

Newton-Anderson at Singular Points

In this paper we develop convergence and acceleration theory for Anderso...
research
07/07/2023

Higher-Order Corrections to Optimisers based on Newton's Method

The Newton, Gauss–Newton and Levenberg–Marquardt methods all use the fir...
research
04/10/2023

Über die Anwendung des Tschebyschew-Verfahrens zum Ausbau des Weierstraß-Kerner-Verfahrens

We extend the Weierstrass-Kerner method by applying the Chebychev method...
research
03/08/2023

The hierarchical Newton's method for numerically stable prioritized dynamic control

This work links optimization approaches from hierarchical least-squares ...
research
11/27/2019

On the choice of initial guesses for the Newton-Raphson algorithm

The initialization of equation-based differential-algebraic system model...
research
02/26/2023

Root finding via local measurement

We consider the problem of numerically identifying roots of a target fun...
research
10/21/2021

Newton-Type Methods For Simultaneous Matrix Diagonalization

This paper proposes a Newton type method to solve numerically the eigenp...

Please sign up or login with your details

Forgot password? Click here to reset