On Fri, Dec 30, 2011 at 11:28:39AM +0100, Andreas Mueller wrote: > It might be that I haven't really understood the meaning of ROC > curves, but I thought it worked like @ogrisel said. > Whatever the correct method to produce a ROC curve > from a linear classifier, I'm pretty sure that using the decision > function is very common in the ML literature.
I agree that the only thing that an ROC curve requires as an input is some arbitrary number, ideally that controls the ratio of false postives vs false negatives. In my experiment, the decision function is indeed a good candidate for that, and thus my remark isn't fully justified. In addition, with a linear-SVM the decision function is really well-behaved in feature space. Thus I should remove my remark that it is a hack. Gael ------------------------------------------------------------------------------ Ridiculously easy VDI. With Citrix VDI-in-a-Box, you don't need a complex infrastructure or vast IT resources to deliver seamless, secure access to virtual desktops. With this all-in-one solution, easily deploy virtual desktops for less than the cost of PCs and save 60% on VDI infrastructure costs. Try it free! http://p.sf.net/sfu/Citrix-VDIinabox _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
