+1 for tackling Bagging first. It's not hard if you are used to the sklearn API.
Alex
On Tue, Oct 16, 2012 at 5:16 PM, Mathieu Blondel wrote:
> Hi,
>
> There used to be a few interesting github PRs, unfortunately closed or
> stalled.
>
> Adaboost:
> https://github.com/scikit-learn/scikit-learn/p
Hi,
There used to be a few interesting github PRs, unfortunately closed or
stalled.
Adaboost:
https://github.com/scikit-learn/scikit-learn/pull/522
Bagging:
https://github.com/scikit-learn/scikit-learn/pull/583
Functional gradient boosting:
https://github.com/scikit-learn/scikit-learn/pull/492
Thanks for the responses. I'm very interested in being able to use base
learners other than decision trees (actually that's what I'm most interested
in, since coming across http://elf-project.sourceforge.net/).
On Oct 16, 2012, at 11:30 AM, Olivier Grisel wrote:
> 2012/10/16 Peter Prettenhofer
2012/10/16 Peter Prettenhofer :
> Hi Siddhant,
>
> currently, we only provide two ensemble techniques: Random/Extra
> Forest and Gradient Boosted Regression Trees - both use decision trees
> as weak learners. Gradient Boosting itself can be used with different
> base learners but AFAIK it hardly is
Hi Siddhant,
currently, we only provide two ensemble techniques: Random/Extra
Forest and Gradient Boosted Regression Trees - both use decision trees
as weak learners. Gradient Boosting itself can be used with different
base learners but AFAIK it hardly is*.
Unfortunately, scikit-learn lacks class
Hi Siddhant.
Depending on how much work you want to invest,
it is possible to implement this, even though it
is currently not supported.
Cheers,
Andy
Am 16.10.2012 10:49, schrieb Gilles Louppe:
> Hi Siddhant,
>
> This is not yet supported unfortunately.
>
> Best,
>
> Gilles
>
> On 15 October 2012
Hi Siddhant,
This is not yet supported unfortunately.
Best,
Gilles
On 15 October 2012 17:50, Siddhant Goel wrote:
> Hi people,
>
> Does scikit-learn support plugging in user defined classifiers in its
> ensemble learning framework? I went through the documentation but could only
> find decis
Hi people,
Does scikit-learn support plugging in user defined classifiers in its ensemble
learning framework? I went through the documentation but could only find
decision trees being used a weak learners. Is it possible to define our own
classifiers and then use an ensemble learning strategy u
