Conditioning (analyzes numerical)

In general, the data of a numerical Problème depend on experimental measurements and are thus entâchées of error.

Conditioning is a quantity which measures the dependence of the solution of a numerical problem compared to the facts of the case, this in order to control the validity of a solution calculated compared to these data.

For example, for a linear System " $Ax=b$" , where the data are the matrix $A$ and the Vecteur second member $b$, conditioning gives a terminal of the relative error made on the solution $x$ when the data $A,\; b$ are disturbed. It can prove that this terminal is very large, so that the error which could result from this makes the solution numerical not exploitable.

Random links:

California | October 25th | Factor of Lorentz | Clathrinida | Corey Robin | Francoise Houdart | Pont_de_Howrah

© 2007-2008 speedlook.com; article text available under the terms of GFDL, from fr.wikipedia.org