Abstract of FKI-156-91

Document-Name:  FKI-156-91 (not available via ftp)
Title:          The Action--Oriented Bucket Brigade
Authors:	Gerhard Weiss
Revision-Date:  August 1991
Category:       Technical Report (Forschungsberichte Künstliche Intelligenz)
Abstract:       Learning in intelligent--agent systems typically works on the 
                principle of action-orientation. In this paper we show that 
                this kind of action--oriented learning can be also realized 
                in classifier systems. In particular, we introduce the action-
                oriented bucket brigade (ABB), a modified and extended version 
                of the standard bucket brigade (BB). Starting from a general 
                overview of current approaches to learning in multi-component 
                systems, first we describe the principle of action-orientation
                and motivate its transfer to learning in classifier systems.
                Next, we introduce the ABB and present its action-oriented
                mechanisms of bidding, firing, and strength adjustment.
                Finally, we focus on the differences between the ABB and 
                the BB.  
Size:           16 pages
Language:       English
ISSN:           0941-6358
Copyright:      The ``Forschungsberichte Künstliche Intelligenz''
                series includes primarily preliminary publications,
                specialized partial results, and supplementary
                material. In the interest of a subsequent final
                publication these reports should not be copied. All
                rights and the responsability for the contents of the
                report are with the authors, which would appreciate
                critical comments.



Gerhard Weiss