Comments on Different techniques for finding best-fit parameters
- Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
A common data analysis problem is to find best-fit parameters through chi-square minimization. Levenberg-Marquardt is an often used system that depends on gradients and converges when successive iterations do not change chi-square more than a specified amount. We point out in cases where the sought-after parameter weakly affects the fit and cases where the overall scale factor is a parameter, that a Golden Search technique can often do better. The Golden Search converges when the best-fit point is within a specified range and that range can be made arbitrarily small. It does not depend on the value of chi-square.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- AC52-06NA25396
- OSTI ID:
- 1136092
- Report Number(s):
- LA-UR-14-24879
- Country of Publication:
- United States
- Language:
- English
Similar Records
LOCO with Constraints and Improved Fitting Technique
Efficient Levenberg-Marquardt minimization of the maximum likelihood estimator for Poisson deviates
Seismic event location : dealing with multi-dimensional uncertainty, model non-linearity and local minima.
Journal Article
·
Thu Jun 18 00:00:00 EDT 2009
· ICFA Beam Dyn.Newslett.44:60-69,2007
·
OSTI ID:1136092
Efficient Levenberg-Marquardt minimization of the maximum likelihood estimator for Poisson deviates
Journal Article
·
Tue Nov 10 00:00:00 EST 2009
· Nature Methods, vol. 7, no. 5, May 1, 2010, pp. 338-339
·
OSTI ID:1136092
Seismic event location : dealing with multi-dimensional uncertainty, model non-linearity and local minima.
Conference
·
Fri Aug 01 00:00:00 EDT 2003
·
OSTI ID:1136092