Extrapolation for Higher Orders of Convergence
Brian Jones
Abstract
Brian Jones
Abstract
A new extrapolation procedure which encompasses Aitken's δ2 process and extends to higher orders of convergence is presented. The new extrapolation yields information as to whether an algorithm is converging or diverging as well as a new approximation to the solution point. The new extrapolation technique is compared with existing higher order extrapolation techniques. The value of the new extrapolation is demonstrated analytically and empirically.
OpenAlex reports 7 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A new extrapolation procedure which encompasses Aitken's δ2 process and extends to higher orders of convergence is presented. The new extrapolation yields information as to whether an algorithm is converging or diverging as well as a new approximation to the solution point. The new extrapolation technique is compared with existing higher order extrapolation techniques. The value of the new extrapolation is demonstrated analytically and empirically.
Key concepts: Extrapolation, Convergence (economics), Applied mathematics, Mathematics, Point (geometry), Mathematical analysis, Geometry, Economics