Rule Generation

  When creating rules one has the choice between creating definite rules or default rules. The definite rules are rules which are correct with respect to all data in the training set. This means that a definite rule has no probability associated with it.

A default or probabilistic rule is a rule which is correct in most cases. This mean that it can describe the most common cases, but that it was not correct for all data in the training set.

In the next subsection we describe the meaning or semantics of definite rules in the form of propositional statements. Then we describe how default rules could reasoned with in a way which is called non-monotonic.

• Semantics of Definite Rules
• Non-Monotonic and Default Reasoning