Zusammenfassung: The algorithm discribed shares with the gradient methods their ability to converge from an initial guess which may be outside the region of convergence of other methods. The algorithm shares with the Taylor series method the ability to close in on the converged values rapidly after the vicinity of the converged values has been reached. Thus, the method combines the best feutures of ith predecessors while avoiding their most serious limitations.
Schlüsselwörter: mathematik, marquardt, nichtlinear, methode, algorithmus