On Sat, Sep 29, 2012 at 10:39:46AM -0700, Ariel Rokem wrote:
> A sneakier (and probably not great) approach would be to have
> ElasticNet itself try calling LinearRegression when alpha is set to 0.
> In a way, that's what a user is asking for by setting alpha to 0 - give
> me the the OLS solution.

So far, we have avoided to do such magic in the scikit. The reasons are
two fold. On the one hand, it is really hard to do it right: should we
put a lower threshold, or do a hard equality on zero? Unless we base our
work on real solutions, and not hacks, we will never be able to hide the 
inner workings and problems of the algorithms from the user. On the other
hand, the more we abstract out the algorithms, the harder we make the to
be control and predicted. A meta algorithms combining many algorithms
into one would have a very large number of meta-parameters, and would end
up being even harder to understand from the standpoint of the user.

Cheers,

Gael

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://ad.doubleclick.net/clk;258768047;13503038;j?
http://info.appdynamics.com/FreeJavaPerformanceDownload.html
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to