Models of Incremental Concept Formation
Abstract
Given a set of observations, humans acquire concepts which organize the observations, and use them to classify future experience. This type of concept formation can occur in the absence of a tutor, and it can take place despite irrelevant and incomplete information. A reasonable model of such concept learning should both be incremental and capable of handling the type of complex experiences that people encounter in the real world. This paper reviews three previous models in incremental concept formation, and then presents Classit, a model that extends the earlier systems. All of the models integrate the process of recognition and learning, and all can be viewed as carrying out searches through the space of possible concept hierarchies. We also present some empirical studies of its behavior under a variety of conditions, in an attempt to show that Classit is a robust concept formation system. Keywords: Conceptual clustering, Discrimination network, Incremental learning, Probabilistic concepts, Concept hierarchy, Concept formation, Category utility, Machine learning.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1988
- Accession Number
- ADA199617
Entities
People
- Douglas Fisher
- John Gennari
- Pat Langley
Organizations
- University of California, Irvine