Prev: problem with mex file and mexw32 file
Next: Locate Minima & Maxima
From: Saari Saari on 6 Jan 2010 04:59
Hello,
I just came forward this newsgroup and found this topic helped me alot in my project since i did the same thing as Gabriel. However, i would like to know how to get standard error for each parameter because before this i tried to use
[stats.se] but it seems did not work for my data. I hope anyone can help me for this question because i really need the value to test the significant for each parameter used. I'm sorry because i am not really good in statistical problem. Thank you.












"thomas " <tjerde_removethispart_(a)stanford.edu> wrote in message <gdl9i2$f15$1(a)fred.mathworks.com>...
> Apologies in advance for my inexpertise on statistical matters.
>
> I'm doing logistic regression using glmfit. I want to compute pseudo-R^2 and AIC (Akaike information criterion), which for a logistic regression seem to require knowing the likelihood estimate for the fitted model. But I can't figure out how to compute this. If I'm not mistaken, glmfit is doing this already in order to iteratively discover the optimized betas, but I can't figure out how to reconstruct that actual likelihood estimate from the outputs it gives me.
>
> Can anyone point me in the right direction? Am I right in thinking that the information I need is already being computed by glmfit, but is not output?
 | 
Pages: 1
Prev: problem with mex file and mexw32 file
Next: Locate Minima & Maxima