Hi Pieter.
Welcome to the project.
There are many open issues. Check the "easy" and "needs contributor" tags.
Also, it might be possible to pick up old pull requests that got stalled.
Best,
Andy
On 08/29/2015 11:47 PM, Pieter de Jong wrote:
Hi all,
I hope to contribute to scikit-learn in the
Hi all,
I hope to contribute to scikit-learn in the near+medium future, and am
following the instructions at http://scikit-learn.org/stable/developers/ ,
leading me email this list. I will continue to follow those instructions
and see where I can contribute; I'm open to any suggestions or pointers
Thanks Andreas,
My actual name is Javier, the name in the address is for anonymity
purposes. Ich spreche trotz ein bisschen Deutsch.
I am getting acquainted with the code and with the algorithms and then I
will decide what to implement.
Thanks for you warm welcome.
2012/9/3 Andreas Mueller
>
Hey Albrecht.
By the way, don't hesitate to ask questions at the issues
if you are unsure on how to address them or how to choose one.
Feedback on the issue-tracker is often near real-time.
Cheers,
Andy
On 09/03/2012 02:09 AM, Analytic Bastard wrote:
Thanks Olivier,
I'll have a look at the URL
This test is purely non-parametric, it is based on the characteristic
function evaluated at the data (OK basically their Fourier transform),
following a characterization of independence by A. Renyi, though its actual
implementation depends on pairwise distances and that is something I must
understa
On Mon, Sep 03, 2012 at 01:54:58AM +0200, Analytic Bastard wrote:
>It would be useful to test whether any classification or regression
>problem can be tractable. After all, if two variables are independent, it
>won't be possible to try to classify or do regression with them.
In my expe
Thanks Olivier,
I'll have a look at the URLs you gave me and comment again at some more
godly hours!
Good night!
2012/9/3 Olivier Grisel
> 2012/9/3 Analytic Bastard :
> > Hi Olivier,
> >
> > Yes that is the coefficient/statistic. Fukumizu, Gretton and coworkers
> also
> > have one based on RKH
2012/9/3 Analytic Bastard :
> Hi Olivier,
>
> Yes that is the coefficient/statistic. Fukumizu, Gretton and coworkers also
> have one based on RKHS covariance operators.
>
> It would be useful to test whether any classification or regression problem
> can be tractable. After all, if two variables ar
O btw, you are right in your application,
Since you are able to capture more non-linearities, a similarity matrix for
time series data or a functional data dataset can be computed with this
non-linear correlation coefficient.
2012/9/3 Analytic Bastard
> Hi Olivier,
>
> Yes that is the coeffici
Hi Olivier,
Yes that is the coefficient/statistic. Fukumizu, Gretton and coworkers also
have one based on RKHS covariance operators.
It would be useful to test whether any classification or regression problem
can be tractable. After all, if two variables are independent, it won't be
possible to t
2012/9/3 Analytic Bastard :
> Hi guys,
>
> My name is Javier and I am a machine learning researcher.
>
> Currently, I am learning Python and found out this excellent package. I
> would like to contribute by implementing any planned feature and by
> implementing some ideas such as a non-linear corre
Hi guys,
My name is Javier and I am a machine learning researcher.
Currently, I am learning Python and found out this excellent package. I
would like to contribute by implementing any planned feature and by
implementing some ideas such as a non-linear correlation coefficient by
SZÉKELY and RIZZO.
12 matches
Mail list logo
