Answer by Prakhar Agarwal for Naive Bayes vs. SVM for classifying text data
Support Vector Machine (SVM) is better at full-length content.Multinomial Naive Bayes (MNB) is better at snippets.MNB is stronger for snippets than for longer documents. While (Ng and Jordan,2002)...
View ArticleAnswer by Horia Coman for Naive Bayes vs. SVM for classifying text data
The biggest difference between the models you're building from a "features" point of view is that Naive Bayes treats them as independent, whereas SVM looks at the interactions between them to a certain...
View ArticleNaive Bayes vs. SVM for classifying text data
I'm working on a problem that involves classifying a large database of texts. The texts are very short (think 3-8 words each) and there are 10-12 categories into which I wish to sort them. For the...
View Article
More Pages to Explore .....
