问题
I am developing a spam filtering application. I need suggestions regarding the hybrid algorithm from Naive Bayes & SVM.(e.g. based on feature vector, probabilities). Any help is appreciated. Can we develop hybrid algorithm from Naive bayes & SVM?
回答1:
Not sure why would you want to merge these two specific methods, but you could use ensemble learning methods for that.
EDIT: based on your comments, it seems you already have two independently trained classifiers, and would like to use them together for classification. One possible method for doing that would be to co-calibrate the classifier results, e.g. via summing and thresholding.
That is: given an input, e.g. an email, both Naive Bayes and SVM generate a real valued result; let's denote these results nb and svm . You could use linear regression to find coefficients c1 and c2 and a threshold t, so that you'd classify an email as spam only if c1 * nb + c2 * svm > t. You could, of course, use more elaborate methods, but with some luck this would give a slightly better performance than each classifier independently.
来源：https://stackoverflow.com/questions/14766404/can-we-obtain-hybrid-algorithm-for-spam-filtering-from-naive-bayes-svm