Thanks Olivier, Andreas. And, again to the text classification module
authors. sklearn rocks!
I think I was quite lucky, but I'm not complaining! :)
My feature set was almost the same as the char and word features that
Andreas used. I found that SVC gave me better performance than LR. And,
some normalizations that I did using a bad words list helped quite a bit as
well. SVC alone would have been enough to win, although I did combine with
RF and it improved the score marginally. These may have been the only
differences. I did not use the datetime field. I had some other features,
some were quite useful, but others not so much, and their effect canceled
out.
On Sat, Sep 22, 2012 at 11:01 AM, Olivier Grisel
<olivier.gri...@ensta.org>wrote:
> 2012/9/22 Andreas Mueller <amuel...@ais.uni-bonn.de>:
> > On 09/22/2012 12:17 PM, Olivier Grisel wrote:
> >> and to Andreas who finished in the 6th position out of 50 final
> submitters.
> >>
> >> This contest was about text classification:
> >>
> >> http://www.kaggle.com/c/detecting-insults-in-social-commentary
> >>
> >> Any feedback on what scikit-learn models where used, which feature
> >> extraction / blending techniques were useful and which were not
> >> working as expected is always appreciated.
> >>
> > *SPAM*
> > My post is here:
> >
> http://peekaboo-vision.blogspot.com/2012/09/recap-of-my-first-kaggle-competition.html
>
> Thanks!
>
> --
> Olivier
> http://twitter.com/ogrisel - http://github.com/ogrisel
>
>
> ------------------------------------------------------------------------------
> How fast is your code?
> 3 out of 4 devs don\\\'t know how their code performs in production.
> Find out how slow your code is with AppDynamics Lite.
> http://ad.doubleclick.net/clk;262219672;13503038;z?
> http://info.appdynamics.com/FreeJavaPerformanceDownload.html
> _______________________________________________
> Scikit-learn-general mailing list
> Scikit-learn-general@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
>
------------------------------------------------------------------------------
How fast is your code?
3 out of 4 devs don\\\'t know how their code performs in production.
Find out how slow your code is with AppDynamics Lite.
http://ad.doubleclick.net/clk;262219672;13503038;z?
http://info.appdynamics.com/FreeJavaPerformanceDownload.html
_______________________________________________
Scikit-learn-general mailing list
Scikit-learn-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
