2014-07-28 23:46 GMT+02:00 Mario Michael Krell <[email protected]>: > I have to somehow contradict. In fact it would be possible to get a > probability but it requires some "work". So it is not easy. > > I my group, we are using a sigmoid fit introduced by Platt to map SVM scores > to probability values. We integrated it in our pySPACE framework, which also > interfaces scikit-learn algorithms. Unfortunately for using the fit together > with kernels additional/separate data is required for training to avoid > over-fitting.
"Easy" was indeed the keyword here. Isotonic calibration is even better than Platt scaling, but hasn't been implemented in scikit-learn master either (there's a PR for it somewhere). ------------------------------------------------------------------------------ Infragistics Professional Build stunning WinForms apps today! Reboot your WinForms applications with our WinForms controls. Build a bridge from your legacy apps to the future. http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
