Re: Which distribution to choose for randomly select calibration data indices?
- From: Xiaoxiao <xiaoxiaoyang@xxxxxxxx>
- Date: Mon, 26 Jan 2009 13:33:06 -0800 (PST)
On Jan 26, 3:15 pm, Peter Perkins
<Peter.PerkinsRemoveT...@xxxxxxxxxxxxx> wrote:
Xiaoxiao wrote:
I have one data, nx1, and I would like to randomly select some of them
as calibration data set, and the left as test data set.
Xiaoxiao, if you have access to the Statistics Toolbox, you can use the CVPARTITION function to create a "holdout" test set and training set. Or you can get fancier and create a K-fold partition. If you don't have the Statistics Toolbox, you probably want to use RANDPERM.
Hope this helps.
Thank you very much for the help, Peter. I have Statistics Toolbox. I
will check the functions you mentioned.
Your suggestions are very helpful, thanks again.
.
- References:
- Prev by Date: Re: how to pick up every column vector from a huge matrix?
- Next by Date: FDATool Get Coefficients
- Previous by thread: Re: Which distribution to choose for randomly select calibration data indices?
- Next by thread: Re: Diff/merge tool
- Index(es):
Relevant Pages
|