Re: SOS: how to fit the negative binomial distribution to a data,requiring the expectd frequency

"epistat" <epistat@xxxxxxxxx> wrote in news:1155970916.167226.55670@
75g2000cwc.googlegroups.com:

Dear friends,
I met an urgent difficulty: I found the S-Plus or Matlab can fit the
NBD( negative binomial distribution ) with maximum likelihood
estimation, but only give the parameter values, How should i get the
corresponding expected frequency and chi-squre,likelihood-values?

Thanks very much!

It might depend on what you meant by "expected frequency" and "chi-square,
likelihood values". And it may depend on the parametrization.

The Wolfram website gives the mean of a NB as k(1-p)/p, whereas Freund says
it is k/p.

I would expect that a GLM function in S-Plus that was estimating a
parameter in a NB model would report a deviance. Was that what you were
looking for when you mentioned chi-square likelihood? Deviance could be
loosely described as -2*log(likelihood) and is asymptotically chi-squared.

See: http://www-stat.stanford.edu/~owen/courses/306a/Splusdiscrete2.pdf
(see p 69 of 256)

--
David Winsemius
.

Relevant Pages

  • Re: Weibull distribution: reference for maximum likelihood inference
    ... estimation of the parameters of the Weibull ... gladly read about unbiased estimation, ... distribution is not very well covered in the ... The large-sample procedure to obtain a confidence interval for h, which mathematically is based on inverting the likelihood ratio test, is to numerically obtain the two-dimensional region ...
    (sci.stat.math)
  • Re: Nonlinear errors in variables models
    ... >My overall approach to the problem would be to combine a likelihood ... the error variances or relative error variances to identify the model. ... name odr2 'simpler ODR estimation, ... run OLS for comparison ...
    (sci.stat.math)
  • MLE - 1st and 2nd derivative of the function
    ... it seems I haven't been doing the maximum likelihood ... This involves millions of calculations. ... maximum likelihood estimation involves the following four ...
    (sci.stat.math)
Loading