[R] GCV UBRE score in GAM models

[Simon Wood]

> I would to know with ranges  of GCV or UBRE values can be considered as
> adequate to consider a GAM as correct

I don't think that there is really an answer to this. It's a bit like
asking what range of values of the residual sum of squares means that a
linear model is correct. I've only seen the GCV score used as a statistic
for judging the relative appropriateness of models, not the `absolute'
appropriateness.

For detecting `incorrectness' of a GAM I'd look at residual plots and in
cases where the scale parameter ought to be one, check that the residuals
are consistent with this.

Simon

_____________________________________________________________________
> Simon Wood simon at stats.gla.ac.uk        www.stats.gla.ac.uk/~simon/
>>  Department of Statistics, University of Glasgow, Glasgow, G12 8QQ
>>>   Direct telephone: (0)141 330 4530          Fax: (0)141 330 4814