On Fri, Apr 13, 2012 at 10:55 PM, Conrad Lee <[email protected]> wrote:
> Thanks for the suggestion of how to change RFE.py to exploit warm_start.
> Should I add this feature to rfe.py and make a pull request? Or is this
> functionality too specialized?
>
+1 for a PR
Do you know of other fast estimators I can use that do check for
> convergence? Or is SGDRegressor in general the fastest? I know speed will
> depend on many factors, what other estimators would you suggest as quite
> quick?
>
Ridge would be a candidate but there is a bit of work to be done to support
warm start. Currently, Ridge supports two solvers: one based on
scipy.linalg.solve in the dense case and another based on
scipy.sparse.linalg.cg in the sparse case. Only the second one can be used
with warm start [1]. You could also add a conjugate-gradient solver for the
dense case too, e.g. with [2].
Mathieu
[1] See option x0 in
http://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.linalg.cg.html#scipy.sparse.linalg.cg
[2] http://docs.scipy.org/doc/scipy/reference/optimize.html
------------------------------------------------------------------------------
For Developers, A Lot Can Happen In A Second.
Boundary is the first to Know...and Tell You.
Monitor Your Applications in Ultra-Fine Resolution. Try it FREE!
http://p.sf.net/sfu/Boundary-d2dvs2
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
