Re: bootstrap and parameter estimation

I would say yes. Take a look at Chapter 7 of Davison & Hinkley's
"Bootstrap Methods and their Applications" (Cambridge, 1997). They
address applying the bootstrap to a variety of model problems such as
survival analysis, generalized linear models, nonlinear & nonparametric
regressions, and other scenarios. Hope this helps.

Jason Clark, PhD
Senior Biometrician, Merck

Leslaw Bieniasz wrote:
> Hello,
>
> As a variation of the previous thread (prior distribution of estimated
> parameters), I have another question:
>
> Assume we perform a nonlinear least square fitting of a model
> to a set of discrete "experimental data" representing points
> on a curve, having certain random errors, the distribution of
> which is not known exactly. The goal is to determine the
> distribution (histogram) of the estimated parameters.
>
> Question: Is this possible to use the bootstrap technique
> for this purpose? If yes, please direct me to the relevant literature.
>
> L.B.
>
> *-------------------------------------------------------------------*
> | Dr. Leslaw Bieniasz, |
> | Institute of Physical Chemistry of the Polish Academy of Sciences,|
> | Department of Electrochemical Oxidation of Gaseous Fuels, |
> | ul. Zagrody 13, 30-318 Cracow, Poland. |
> | tel./fax: +48 (12) 266-03-41 |
> | E-mail: nbbienia@xxxxxxxxxxxxx |
> *-------------------------------------------------------------------*
> | Interested in Computational Electrochemistry? |
> | Visit my web site: http://www.cyf-kr.edu.pl/~nbbienia |
> *-------------------------------------------------------------------*

.

Relevant Pages

  • Re: The illusion of the bootstrap technique
    ... A properly specified parametric technique (such as your normal ... If you *knew* that the underlying distribution for the data was normal, ... You will beat the bootstrap and all is peachy. ... and through many replications. ...
    (sci.stat.math)
  • Re: The BOOTSTRAP trap
    ... samples get a t-statistic as big or bigger than the t-stat from the ... would be a function of the spread of each distribution and the distance ... the expectation of the differences of the bootstrap sample and the ... population parameter from a sample is just as much chasing our tail as ...
    (sci.stat.math)
  • Re: Kurtosis, skewness and bootstrapping
    ... > sample from S1 would be the average of the kurtosis values for every ... > possible bootstrap sample. ... > the hypothesis that the kurtosis values of the initial populations ... > would give us a better and better estimate of the distribution of d. ...
    (sci.stat.math)
  • bootstrap and parameter estimation
    ... As a variation of the previous thread (prior distribution of estimated ... Assume we perform a nonlinear least square fitting of a model ...
    (sci.stat.math)