3 ms·
Naive Bayes classifiers are simple, (relatively) easy to understand and fairly straight-forward to code. They are also fairly robust and work well for a wide v
by nl 16y ago
Naive Bayes classifiers are simple, (relatively) easy to understand and fairly straight-forward to code.
They are also fairly robust and work well for a wide variety of problem sets.
Other techniques sometimes offer some improvement, but often don't. Generally a Bayesian classifier is a good place to start.