Conditional Independence

• A Naive Bayes classifier assumes that the attribute values are independent of each other given the class. Normally distributed: many statistical methods assume that data is normally distributed.