On Mon, Sep 24, 2012 at 06:34:36PM +0200, Sheila the angel wrote:
> 1. I am trying to understand how exactly this probability is calculated. The
> document says "probability model is created using cross validation"

Using Platt scaling that is calibrated by cross-validation.

> So I think to calculate probability it do cross validation N times and counts
> how many time data point is in right class and calculates this probability. Is
> it this description right? If so how many time this cross validation is done
> (The value of N) .  

I think that it's a leave one out, but you should check libsvm's
documentation.

> 2. Again from documents "it will produce meaningless results on very small
> datasets"... how small is very small?

I am sorry, I do not know.

HTH,

Gaël

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to