scikit learn - Ensembling with dynamic weights -
i wondering if possible use dynamic weights in sklearn's votingclassifier. overall have 3 labels 0 = other, 1 = spam, 2 = emotion. dynamic weights mean following:
i have 2 classifiers. first 1 random forest performs best on spam detection. other 1 cnn superior topic detection (good distinction between other , emotion). votingclassifier gives higher weight rf when assigns label "spam/1".
is votingclassifier right way go?
best regards,
stefan
Comments
Post a Comment