A Comparison Between Naïve Bayes Classifier and EM Algorithm

in this work we would study and compare two algorithms from Bayesian reasoning family. Bayesian reasoning is based onthe assumption that the quantities of interest are governed by probability distributions and that optimal decisions can be made byreasoning about these probabilities together with observed data. In this study we would try to compare EM algorithm and Naïve Bayesclassifier by detecting their weaknesses and present good solutions for them. Some of these weaknesses expressed in previous workssuch as initializing the parameters in Naïve Bayes classifier. In these situations we tried to present a more efficient solution than beforeworks. Some of these weaknesses expressed for first time in this paper such as the problem of means equality of two or more clusters.

Keywords:

EM algorithm , Naïve Bayes classifier