SIAM Journal on Numerical Analysis · 1974 · 93 citations · 4 references
ComputingComputer ScienceAlgorithmsComputational MechanicsTheory Of ComputationApproximation AlgorithmsApproximation MethodsConvergence AnalysisNumerical AlgorithmsNumerical AnalysisRoot FindingConvergenceNewton ConvergencePolynomial Root FindingRoot Finding AlgorithmsSimultaneous Approximation
This note gives an analysis of the order of convergence of some modified Newton methods. The modifications we are concerned with are well-known methods—a total-step method and a single-step methods—for refining all roots of an nth-degree polynomial simultaneously. It is shown that for the single-step method the R-order of convergence, used by Ortega and Rheinboldt in [6], is at least $2 + \sigma _n > 3$, where $\sigma _n > 1$ is the unique positive root of the polynomial $p_n (\sigma ) = \sigma ^n - \sigma - 2$.
4
James H. Bramble, Richard S. Varga · Mathematics of Computation · 1963 · 4.2K citations
A modified Newton method for polynomials
Louis W. Ehrlich · Communications of the ACM · 1967 · 220 citations · Full text
A posteriori error bounds for the zeros of polynomials
W. B�rsch-Supan · Numerische Mathematik · 1963 · 66 citations