Naive Bayes

Context

• The Naive Bayes network is perhaps the best-known type of Bayesian network.
• In the field of computer science, it is a widely used model.
• A Naive Bayes network has a predefined structure in which the Target Node is the parent of all the other nodes.
• This structure implies that the Target Node is the cause of all the other nodes and that the knowledge of its value makes each node independent of the others.
• Despite these strong assumptions, which are unjustifiable in most cases, the small number of probabilities to estimate makes this structure very robust, with a short learning time.
• Note that no structural learning has to take place: the arc structure is given, and only the parameters, i.e., the Conditional Probability Tables have to be estimated.