I'm actually working on a Twitter sentiment analysis project in Python. The chosen classifiers were Linear support vector machines, Max entropy and Naive Bayes. Is it possible that the accuracy of the NB classifier is higher than linear SVM of 0.001?
-
$\begingroup$ Yes, that is possible. $\endgroup$ – Marc Claesen Jan 30 '15 at 13:57
-
$\begingroup$ @MarcClaesen What are the factors who influence this? $\endgroup$ – JackLametta Feb 1 '15 at 17:10