Extends the traditional classification of errors so that the error of the method (truncation error) and the numerical error are subdivided into four classes: the approximation, the perturbation, the algorithm and the rounding error. This new subdivision of errors results in error estimates for a num
Error Analysis in Numerical Processes
โ Scribed by Solomon G. Mikhlin
- Publisher
- Wiley
- Year
- 1991
- Tongue
- English
- Leaves
- 288
- Series
- Pure and Applied Mathematics
- Edition
- 1
- Category
- Library
No coin nor oath required. For personal study only.
โฆ Synopsis
Extends the traditional classification of errors so that the error of the method (truncation error) and the numerical error are subdivided into four classes: the approximation, the perturbation, the algorithm and the rounding error. This new subdivision of errors results in error estimates for a number of linear and nonlinear problems in numerical analysis. Obtained here are new results--errors in the conjugate direction method--as well as known results, such as errors in Gaussian elimination. Also presented are a posteriori error estimates, such as those derived for, and often by means of, the computed approximate solution.
โฆ Subjects
Mathematical Analysis;Mathematics;Science & Math;Mathematics;Algebra & Trigonometry;Calculus;Geometry;Statistics;Science & Mathematics;New, Used & Rental Textbooks;Specialty Boutique
๐ SIMILAR VOLUMES
Extends the traditional classification of errors so that the error of the method (truncation error) and the numerical error are subdivided into four classes: the approximation, the perturbation, the algorithm and the rounding error. This new subdivision of errors results in error estimates for a num
<p>Owing to the developments and applications of computer science, maยญ thematicians began to take a serious interest in the applications of number theory to numerical analysis about twenty years ago. The progress achieved has been both important practically as well as satisfactory from the theoretic