On Sun, Mar 24, 2013 at 11:41 AM, Mathieu Blondel wrote:
> If you use a recent version of scikit-learn, you should be able to do this:
> ovr = OneVsRestClassifier(SGDClassifier(loss="log"))
> ovr.fit(X_train, y_train)
> prob = ovr.predict_proba(X_test)
Ah, right, you want to use partial_fit so m
On Sun, Mar 24, 2013 at 7:37 AM, ShNaYkHs ShNaYkHs wrote:
> Well, I need a classifier in that can support partial_fit() and can give a
> probability for prediction proba_predict() (in case of multi-class, not only
> binary). PassiveAggressiveClassifier and Perceptron have no proba_predict()
> meth
Well, I need a classifier in that can support partial_fit() and can give a
probability for prediction proba_predict() (in case of multi-class, not
only binary). PassiveAggressiveClassifier and Perceptron have no
proba_predict() method, and SGDClassifier have a proba_predict() method but
supported o
2013/3/21 Olivier Grisel :
>> Also, I'm not sure what the state of the MLP is. Lars is working on it
>> and Temporaer (aka Hannes) has been
>> trying to help out there.
>
> Lars would you like to mentor a GSoC on this? Or do you prefer to
> finish your PR independently of a GSoC?
I'm sorry, but I
