Prev: How to control Matlab's printed number field width?
Next: two way anova repeated measure
From: Euh on 6 Apr 2010 18:19
I'm trying to optimize the parameters of a model so as to minimize the
difference between measured and predicted values.
I'm able to compute the confidence interval for each parameter.


1) Some parameters have large confidence intervals (basically related
to the upper and lower bounds provided as constraints)

2) Removing some parameters does not significantly affect the sum
squared of the residuals

1) is telling me which parameters are observable (for the given set of
measurements) right ?

Do 1) and 2) necessarily go hand in hand ?

In other wordz, what is the best justification for model reduction ?



 | 
Pages: 1
Prev: How to control Matlab's printed number field width?
Next: two way anova repeated measure