2
$\begingroup$

If I use an ensemble, which consists of four classifiers, in order to classify my data into three classes. Further, suppose I use the one versus one strategy. My question is: How to fuse the probabilities of the ensemble? (That is, since we have two probabilities for each class in each classifier.)

$\endgroup$
  • $\begingroup$ +1 Great question. A majority vote is typically held to make the final prediction. I'd say that you'd average all the probabilities for each respective class. Also - I do not think it is appropriate to use the term ensemble here, as it may be confused with ensemble learning. $\endgroup$ – user46925 Mar 24 '16 at 2:07
  • $\begingroup$ Thank Mr.Gung for the answer, however I built an ensemble model which consist of four classifiers, and each classifier has to use one versus one in order to classify the three classes, then I have to use the ( as you say the majority voting) . can I use the algebra methods like the product rule which multiply the average of each probability in the ensemble model. $\endgroup$ – Dheeb bashish Mar 25 '16 at 2:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.