Naive Bayesian Classiier Committees

The naive Bayesian classiier provides a very simple yet surprisingly accurate technique for machine learning. Some researchers have examined extensions to the naive Bayesian classiier that seek to further improve the accuracy. For example, a naive Bayesian tree approach generates a decision tree with one naive Bayesian classiier at each leaf. Another example is a constructive Bayesian classiier that eliminates attributes and constructs new attributes using Cartesian products of existing attributes. This paper proposes a simple, but eeective approach for the same purpose. It generates a naive Bayesian classiier committee for a given classiication task. Each member of the committee is a naive Bayesian classiier based on a subset of all the attributes available for the task. During the classiication stage, the committee members vote to predict classes. Experiments across a wide variety of natural domains show that this method signiicantly increases the prediction accuracy of the naive Bayesian classiier on average. It performs better than the two approaches mentioned above in terms of higher prediction accuracy.