home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.ai:4716 comp.ai.edu:589
- Newsgroups: comp.ai,comp.ai.edu
- Path: sparky!uunet!haven.umd.edu!news.umbc.edu!gmuvax2!twallace
- From: twallace@gmuvax2.gmu.edu (Todd Wallace)
- Subject: Re: Looking for PC based expert systems.
- Message-ID: <1992Dec30.165211.28308@gmuvax2.gmu.edu>
- Keywords: general, learning
- Organization: George Mason University, Fairfax, Va.
- References: <71720001@hplvec.LVLD.HP.COM> <1g263tINNh1j@sol.deakin.OZ.AU>
- Distribution: inet
- Date: Wed, 30 Dec 1992 16:52:11 GMT
- Lines: 30
-
- I am working on a model of a generalized learning paradigm to be
- modelled in an object-oriented environment (C++). Ideally, I
- would like to end up with a module that you "plug in" to a
- program, hooking up its inputs and outputs, set rewards and
- goals, and off it would go, figuring out its environment as it
- went.
-
- I am approaching this problem from the point of view of
- biological system modelling, something like if you were to
- design a Turing machine or Life game "cell" that is reinforced
- for performing certain actions in its environment. For
- instance, you could theoretically find a Turing machine that
- would sort letters introduced into the left edge of its
- environment and transport them to the right edge of its
- environment. Each Turing machine could form a link in a whole
- assembly-line type "factory" that would eventually make a
- working program.
-
- I don't want to reinvent the wheel, though, so I'd like to start
- out by compiling a bibliography of related research. However, I
- am handicapped because I don't know what this line of AI
- research would be called. Does anyone out there in AI land
- know?
- Thanks for your help.
- ---------------------
- Todd Wallace
- twallace@gmuvax2.gmu.edu
-
-
-
-