Newsgroups: comp.ai.neural-nets,comp.ai.fuzzy,comp.ai
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!hookup!swrinde!howland.reston.ans.net!spool.mu.edu!umn.edu!peca!hougen
From: hougen@peca.cs.umn.edu (Dean Hougen)
Subject: Re: ? NN Solved the triple inverted pendulum yet?
Message-ID: <D5I0qL.E1n@news.cis.umn.edu>
Summary: I don't think so.
Keywords: pole-balancing
Sender: news@news.cis.umn.edu (Usenet News Administration)
Nntp-Posting-Host: peca.cs.umn.edu
Organization: University of Minnesota, Minneapolis, CSci dept.
References: <3jq0pb$o7l@tribune.usask.ca>
Date: Wed, 15 Mar 1995 20:25:11 GMT
Lines: 23
Xref: glinda.oz.cs.cmu.edu comp.ai.neural-nets:22736 comp.ai.fuzzy:4216 comp.ai:28159

In article <3jq0pb$o7l@tribune.usask.ca> choy@cs.usask.ca (Henry Choy) writes:
>Has a neural net been successfully applied to the triple inverted
>pendulum? A fuzzy logic solution has been claimed.

Well, if a NN has solved the triple inverted pendulum problem, it would
be news to me.  I've seen just about every NN inverted pendulum paper
that's been published, so unless it was very recent or very obscure, I
don't think its been done.

Then again, I haven't seen the fuzzy-logic triple inverted pendulum
solution (although I really don't follow fuzzy-logic writings).  Tell
me more about it.  Where was the claim published?  Does the system
*learn* to balance the pole (cool), or did someone simply design a
fuzzy-logic solution (yawn)?  Does the system partition the state
space (really cool), or did the researcher do that (okay, but nearly
as impressive)?  Anyway, the citation is the important question.


>        Henry Choy                       "Math class is hard" - Barbie

Dean Hougen
--
"Its a hard logic to follow when the girls get lost."  - Talking Heads
