CMU Artificial Intelligence Repository
NEITHER: NEITHER Theory Refinement System
areas/learning/systems/utexas/neither/
NEITHER, which stands for "New EITHER", is an extension of the EITHER
theory refinement system. Theory refinement systems are used to modify
a set of rules which are incomplete or incorrect. Given a set of
examples, the rules are refined until the rule base is consistent with
the examples. One example of the use of theory refinement is the
knowledge acquisition phase of expert system construction.
Specifically, it may be relatively simple to develop a set of rules to
cover most of a problem but difficult to generate a complete solution.
Often, additional examples can be generated easily and given with the
partially correct theory to a system like NEITHER. The result is an
improved set of rules.
Origin:
cs.utexas.edu:/pub/mooney/
Requires: Common Lisp
Copying: Copyright (c) 1991 by Raymond J. Mooney
Use permitted for educational and research purposes only.
Users are requested, but not required, to inform Raymond
J. Mooney of any noteworthy uses of this software.
Please see the file readme.txt for details.
CD-ROM: Prime Time Freeware for AI, Issue 1-1
Contact: Raymond J. Mooney
Machine Learning Group
Department of Computer Sciences
The University of Texas at Austin
Keywords:
Authors!Mooney, Lisp!Code, Machine Learning, NEITHER,
Teaching Materials, Theory Refinement
References: ?
Last Web update on Mon Feb 13 10:24:35 1995
AI.Repository@cs.cmu.edu