| Submitted for publication in Machine Learning Journal, 1991 LEARNING TWO-TIERED DESCRIPTIONS (2007) | |||||||||||||||||
Abstract | |||||||||||||||||
| email: jzhang @ gmuvax2.gmu.edu Machine learning research has so far been primarily concerned with learning crisp concepts, that is concepts that are well-defined and context-independent. Most real-life concepts, however, are flexible, as their meaning is not precise and can change with the context in which they are used. To extend the applications of machine learning, it is therefore important to develop methods for learning flexible concepts. This paper describes such a method, which is based on a two-tiered concept representation. In the two-tiered representation, the first tier, called the Base Concept Representation, contains an explicit description of core concept properties, and the second tier, called the Inferential Concept Interpretation, defines allowable modifications of the explicit meaning and its dependence on the context of discourse. The method generates a Base Concept Representation by first creating a complete and consistent concept description from supplied training examples, and then optimizing the description according to a general description quality criterion. The complete and consistent description is obtained by applying the AQ inductive learning methodology. The optin'fization process is done by a best-f'rrst | |||||||||||||||||
Publication details | |||||||||||||||||
| |||||||||||||||||