choosing a classifier really depends on application.

You need to perform tests for generalisation abilities of the classifiers (e.g. cross-validation tests, leave-one-out, McNemar test, etc.). For example, the book: Introduction to Machine Learning (by Ethem Alpaydinm), provides a helpful introduction to topic (Chapter: Design and Analysis of Machine Learning Experiments). 

you can try for naive bayes classifier.

Hi,

If I understand properly your question, all of them could give you really good results to classify data in different sets, but the selection of the best one completly depends on the system where you are going to apply the neural network. In other words, to answer your question is necessary to apply the algorithms directly in your system.

Cheers.

José.