From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!jupiter!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!uunet!trwacs!erwin Tue Nov 26 12:32:19 EST 1991
Article 1579 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca rec.arts.books:10622 sci.philosophy.tech:1110 comp.ai.philosophy:1579
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!jupiter!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!uunet!trwacs!erwin
>From: erwin@trwacs.UUCP (Harry Erwin)
Newsgroups: rec.arts.books,sci.philosophy.tech,comp.ai.philosophy
Subject: Re: Zeleny (was Re: Searle
Message-ID: <441@trwacs.UUCP>
Date: 25 Nov 91 13:52:30 GMT
References: <1991Nov14.223348.4076@milton.u.washington.edu> <MATT.91Nov24000158@physics.berkeley.edu> <1991Nov24.195230.5843@husc3.harvard.edu> <1991Nov24.224724.2149@arizona.edu>
Followup-To: rec.arts.books
Distribution: world,local
Organization: TRW Systems Division, Fairfax VA
Lines: 13

The assumption that humans have a mystic power to "induct" is interesting.
This summer, my son tried to train back-prop nets on chaotic data to see
how good they were at predicting chaotic time series _at arbitrary
points_. He learned that the nets were no good unless they had experience
in the subset of the state space that he was testing them on. No magical
power exists for the nets to make predictions without experience. That's
consistent with my experience teaching math. People "induct" because the
brain is good at learning inductive patterns.

Cheers,
-- 
Harry Erwin
Internet: erwin@trwacs.fp.trw.com


