Naive Bayes Classifier

• results comparable to ANN and decision trees in some domains
• Each instance is described by a conjunction of attribute values and the target value can take any value from a set . A set of training instances are provided and a new instance is presented and the learner is asked to predict the target value.
• 
• is estimated by the frequency of each target value in the training data
• cannot use frequency for unless we have a very,very large set of training data to get a reliable estimate
• naive Bayes assumes attribute values are conditionally independent given the target value -
• Naive Bayes classifier: , where denotes the target value
• can be estimated by frequency
• when conditional independence assumption is satisfied the naive Bayes classification is a MAP classification
• naive Naive Bayes entails no search!!