I agree, but instead of changing the return results. We could have a
routine to return the illustration of the one-vs-one classifier used. At
least that is what the e1071 package in R using:
setosa/versicolor setosa/virginica versicolor/virginica
1 1.196152 1.091460 0.6705626
2 1.064621 1.056332 0.8479934
3 1.180842 1.074534 0.6436474
4 1.110699 1.053143 0.6778595
On Mon, Mar 26, 2012 at 10:08 PM, Mathieu Blondel <[email protected]>wrote:
> Since SVC.decision_function is always confusing everyone, instead of
> returning an array of shape [n_samples, n_classes * (n_classes - 1) / 2],
> we could return an array of shape [n_samples, n_classes] with the votes
> obtained by each class. Then an argmax() on this array would lead to the
> same results as predict().
>
> Mathieu
>
>
> ------------------------------------------------------------------------------
> This SF email is sponsosred by:
> Try Windows Azure free for 90 days Click Here
> http://p.sf.net/sfu/sfd2d-msazure
> _______________________________________________
> Scikit-learn-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
>
--
Best Wishes
--------------------------------------------
Meng Xinfan（蒙新泛）
Institute of Computational Linguistics
Department of Computer Science & Technology
School of Electronic Engineering & Computer Science
Peking University
Beijing, 100871
China
------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
