Classifiers
Classifiers typically used when we are interested in predicting the value of one single variable (the class) on the basis of the values of the other variables in the database (the attributes) and we want to learn the probabilistic relationships between class and attributes from a database of cases.
Naive Bayes Classifiers
Naive Bayes Classifiers are among the most commonly used methods for supervised classification. The characteristic assumption of Naive Bayes Classifiers is that the attributes are conditionally independent given the class variable.
A Naive Bayes Classifier can be represent as a Bayesian Network in which the class variable is the common parent node of all attributes. A Bayesian Network representing a Naive Bayes Classifier looks like the following:
In this case, the variable Play represent the class variable and all the other variables represent the attributes. The learning task reduces to the definition of the model of independence underlying the Naive Bayes Classifier and to the estimation of the conditional probability distributions of the resulting network.
Once a classifier has been trained on the data, it can be used to predict the class value of a set of attributes values, by propagating the attributed values as evidence of any Bayesian Network.