This book presents a theory of learning from examples called Nested Generalized Exemplar (NGE) theory, and demonstrates its importance with empirical results in several domains. Nested Generalized Exemplar theory is a variation of a learning model called exemplar-based learning, which was originally proposed as a model of human learning by Medin and Schaffer . In the simplest form of exemplar-based learning, every example is stored in memory verbatim, with no change of representation. The set of examples that accumulate over time form category definitions; for example, the set of all chairs that a person has seen forms that person’s definition of “chair.” An example is normally defined as a vector of features, with values for each feature, plus a label which represents the category of the example.
Unable to display preview. Download preview PDF.