MathGroup Archive 2000

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Levenberg-Marquardt ?

  • To: mathgroup at
  • Subject: [mg23770] Re: Levenberg-Marquardt ?
  • From: Jens-Peer Kuska <kuska at>
  • Date: Sat, 10 Jun 2000 02:59:16 -0400 (EDT)
  • Organization: Universitaet Leipzig
  • References: <8hfddu$>
  • Sender: owner-wri-mathgroup at



has it, as one of the possible build in methods.

Here is the text from the Help-Notebook:

"NonlinearFit and NonlinearRegress use FindMinimum to find parameter
estimates, thus both functions accept the FindMinimum options
AccuracyGoal, Compiled, Gradient, MaxIterations, Method, PrecisionGoal,
and WorkingPrecision. The Method option allows you to choose between
algorithms for performing the minimization of the
\!\(TraditionalForm\`\[Chi]\^2\) merit function. The LevenbergMarquardt
method gradually shifts the search for the minimum of
\!\(TraditionalForm\`\[Chi]\^2\) from steepest descent to quadratic
minimization. Other possible settings for Method are Gradient (steepest
descent), Newton, QuasiNewton, and Automatic. The Automatic method does
linear fitting for linear models and LevenbergMarquardt search for
nonlinear models, thus the meaning of the automatic method differs
between NonlinearFit (or NonlinearRegress) and FindMinimum." 


Alastair McLean wrote:
> Has the Levenberg-Marquard method of non-linear least
> squares minimisation been implemented as a Mathematica
> function or package yet ?
> Alastair McLean
> mclean at

  • Prev by Date: Re: joining lists.
  • Next by Date: Re: system of nonlinear ODE
  • Previous by thread: Re: Levenberg-Marquardt ?
  • Next by thread: Re: Levenberg-Marquardt ?