Thanks for your information, I've understood the R^2.
At 2015-05-06 03:17:50, "Michael Eickenberg" <michael.eickenb...@gmail.com>
wrote:
Under gaussian (and most other) noise and a nonexplicative model the
distribution of r^2 has a negative mode.
As a consequence, sometimes an r^2 score of 0 already implies significant
predictive capacity... happy publishing!
On Tuesday, May 5, 2015, Chris Holdgraf <choldg...@berkeley.edu> wrote:
To my knowledge, R2 is basically a comparison of your model fit, to the model
that is fit by simply drawing a line through the mean of your output variable.
To that extent, if your fit model does worse than the "mean fit", then R2 will
be negative. E.g., check out:
http://randomanalyses.blogspot.com/2011/11/coefficient-of-determination-r2.html
------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
Scikit-learn-general mailing list
Scikit-learn-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
