Small Errors Imply Large Instabilities
Numerical Analysts and scientists working in applications often observe that once they improve their techniques to get a better accuracy, some instability creeps in through the back door. This paper shows for a large class of numerical methods that such a Trade-off Principle between error and stability is unavoidable. The setting is confined to recovery of functions from data, but it includes solving differential equations by writing such methods as a recovery of functions under constraints imposed by differential operators and boundary values. It is shown in particular that Kansa's Unsymmetric Collocation Method sacrifices accuracy for stability, when compared to symmetric collocation.
READ FULL TEXT