On Fri, Dec 2, 2011 at 1:00 PM, Vincent Dubourg
<[email protected]> wrote:
> On 02/12/2011 18:19, Alexandre Passos wrote:
>> On Fri, Dec 2, 2011 at 12:02, James Bergstra<[email protected]>  
>> wrote:
>>> On Tue, Nov 29, 2011 at 5:24 PM, Olivier Grisel
>>> <[email protected]>  wrote:
>>>> That makes sense. Fortunately we don't have an API to compute the
>>>> expected variance of a prediction :)
>> So what does the eval_MSE option do?
> Indeed, AlexP is right. The eval_MSE kwarg bool does exactly what you
> want and reuse intermediate variables to make the computation of the
> prediction's mean and variance optimal (as it is done in DACE for Matlab
> on which this implementation is based). When eval_MSE is True, the
> predictor returns a tuple of 2 arrays (one for the mean values and the
> other for the variance's) that you can split right away. See the
> following examples:
> http://scikit-learn.org/stable/auto_examples/gaussian_process/plot_gp_regression.html
> http://scikit-learn.org/stable/auto_examples/gaussian_process/plot_gp_probabilistic_classification_after_regression.html
>
> Cheers,
>

Cool, good to know!

On the name though --- "eval_MSE" is a nonstandard term for "variance"
no? MSE usually refers to a loss criterion, for comparing predictions
with targets.

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to