Copy Link
Add to Bookmark
Report
AIList Digest Volume 8 Issue 063
AIList Digest Tuesday, 23 Aug 1988 Volume 8 : Issue 63
Queries and Responses:
Category Theory in AI
Camera Stabilization
Speech rec. using neural nets
Sigmoid function
MACSYMA Availability
ELIZA
----------------------------------------------------------------------
Date: 18 Aug 88 10:23:55 GMT
From: mcvax!ukc!strath-cs!glasgow!jack@uunet.uu.net (Jack Campin)
Subject: Re: Category Theory in AI
geddis@atr-la.atr.junet (Donald F. Geddis) wrote:
>>dpb@philabs.philips.com (Paul Benjamin) writes:
>> Some of us here at Philips Laboratories are using universal
>> algebra, and more particularly category theory, to formalize
>> concepts in the areas of representation, inference and
>> learning.
>I'm familiar with those areas of AI, but not with category theory (or
>universal algebra, for that matter). Can anyone give a short summary for the
>layman of those two mathematical topics? And perhaps a pointer as to how
>they might be useful in formalizing certain AI concepts. Thanks!
A short summary is tricky without knowing your mathematical background and
maybe impossible for a real honest-to-goodness layman. A good book to start
with is Herrlich and Strecker's, but if you don't know what a group is, forget
it. Arbib and Manes' "Arrows, Structures and Functors" is also OK, but mainly
applies it to automata theory (not a booming enterprise these days).
Category theory generalizes the notions of "set" and "function", or more
generally "mathematical structure" and "mapping that preserves that structure"
(where the structures might be, say, n-dimensional Euclidean spaces, and the
mappings projections, embeddings and other distance-preserving functions).
Its aim is to describe classes of mathematical object (groups, topological
spaces, partially ordered sets, ...) by looking at the maps between them, and
then to describe relationships between these classes. It captures a lot of
otherwise indescribable mathematical notions of "unique" or "natural" objects
or maps in a class (the empty set, Descartes' construction of the Euclidean
plane as the "product" of two lines, the class of all possible strings in an
alphabet, ...).
The major application of it to computer science so far is in the semantics
of higher-order polymorphic type systems (which can't be described in set
theory). David Rydeheard and Rod Burstall have just published a book
"Computational Category Theory" that describes categorical constructions
algorithmically (in Standard ML) and has a useful bibliography.
But a lot of computer science literature that uses category theory does not
do so in an essential way; the commutative diagrams are just there to give
the authors some mathematical street cred.
I can't imagine what category theory has to contribute to knowledge
representation (though I can just about imagine it helping to describe
neural nets in a more abstract way). Can the philabs people say more
about what they're up to?
--
ARPA: jack%cs.glasgow.ac.uk@nss.cs.ucl.ac.uk USENET: jack@cs.glasgow.uucp
JANET:jack@uk.ac.glasgow.cs useBANGnet: ...mcvax!ukc!cs.glasgow.ac.uk!jack
Mail: Jack Campin, Computing Science Dept., Glasgow Univ., 17 Lilybank Gardens,
Glasgow G12 8QQ, SCOTLAND work 041 339 8855 x 6045; home 041 556 1878
------------------------------
Date: 18 Aug 88 16:48:27 GMT
From: pacbell!hoptoad!dasys1!step!perl@ames.arpa (Robert Perlberg)
Subject: Re: Camera Stabilization
In a previous article, John B. Nagle writes:
> Panasonic showed a gyroscopically stablized consumer-grade camcorder
> at the Consumer Electronics Show this summer. It should be available at
> Japanese retailers by the end of the year. This may be a promising approach.
The Panasonic Steadi-Cam has no gyros. The lens and image sensor are
mounted on a gimbaled platform along with pitch and yaw sensors which
drive motors which move the platform to compensate for camera body
movement.
Robert Perlberg
Dean Witter Reynolds Inc., New York
phri!{dasys1 | philabs | manhat}!step!perl
-- "I am not a language ... I am a free man!"
------------------------------
Date: 19 Aug 88 18:05:25 GMT
From: att!chinet!mcdchg!clyde!watmath!watvlsi!watale!dixit@bloom-beaco
n.mit.edu (Nibha Dixit)
Subject: Speech rec. using neural nets
Is anyody out there looking at speech recognition using neural
networks? There has been some amount of work done in pattern
recognition for images, but is there anything specific being done
about speech?
--
Nibha Dixit (U of Waterloo, Waterloo, Ont.)
...!watmath!watale!dixit or dixit@watale.waterloo.cdn
dixit@watale.waterloo.edu or dixit@watale.waterloo.bitnet
------------------------------
Date: Sat, 20 Aug 88 15:06:38 -0200
From: Antti Ylikoski <ayl%hutds.hut.fi%FINGATE.BITNET@MITVMA.MIT.EDU>
Subject: Sigmoid function
One way to build a circuit that produces the true sigmoid function
would be to store the value-argument pairs into a ROM and use the
following circuit:
f
|-----------------| |------| |-----------------|
input ---> | A / D converter |->| ROM |->| D / A converter | ---> output
|-----------------| |------| |-----------------|
Antti (Andy) Ylikoski
Helsinki University of Technology (I think a better translation would be
Helsinki Institute of technology)
Digital Systems Laboratory
YLIKOSKI@FINFUN.BITNET
OPMVAX::YLIKOSKI (DECnet)
mcvax!hutds!ayl (UUCP)
------------------------------
Date: Sat, 20 Aug 88 23:41 EDT
From: Nick Papadakis <AILIST-REQUEST@AI.AI.MIT.EDU>
Subject: MACSYMA Availability
If the messages I have received since my posting in AIList V8
#35 are any gauge, there seems to be a fair amount of misinformation on
the subject of MACSYMA and where to get it.
The following is a summary of the best information I have been
able to garner via numerous telephone conversations (I had requested
hardcopy, but am tired of waiting for it to arrive).
* * *
The original MACSYMA code is owned by MIT, which has granted
licenses to distribute it to three other organizations.
SYMBOLICS: Runs on Apollo, Sun, Symbolics, and all VAXes.
Licenses range from about $5K to $15K (US prices, for commercial
customers) Probably the most sophisticated version, many
enhancements.
Source code is NOT provided.
Call 1-800-622-7962, (in Mass. (617) 621-7770)
or email petti@ALLEGHENY.SCRC.Symbolics.COM
NESC (National Energy Software Center): Referred to as 'DOE MACSYMA'.
Runs on Alliant, Celerity, Data-General, Encore, LMI Lambda,
Sun, Symbolics, TI Explorer, VAX.
About $2K to $3K for non-subscribers (subscribers get 2 programs
free, subscriptions are $2.5K to $3.5K)
Source code IS provided.
Call Margaret Butler (312) 972-7250
[As of March 87, an improved version of DOE MACSYMA for the
TI Explorer (including the SHARE libraries), was available to any
NESC licensee from Hyde%NGSTL1@TI-CSL.CSNET@RELAY.CS.NET]
INTERMATH - This startup company's product is still in the works, but they
are interested in talking to people who "might want to embed some
portion of MACSYMA's functionality in another system".
Call (617) 868-4510
* * *
Now for the distressing part.
The MIT patent office states that DOE (via NESC) is only
permitted to distribute MACSYMA to government agencies, contractors, and
grantees. NESC says that is completely untrue, and that they will
continue to distribute to commercial customers. Symbolics has
trademarked the name 'MACSYMA'. DOE claims that "it wasn't theirs to
trademark."
I'm not sure I want to know precisely what causes such a massive
failure of communication.
I am _quite_ sure that I do *not* wish to receive (and will not
post) any more messages pointing out that the version being distributed
by a certain company was 'the only licensed version' and that all others
were 'bootleg'. This list exists to inform the AI community, and not to
serve any commercial interest.
It is clear that the various versions available have varying
degrees of enhancement and support. The informed customer will take
this into account when making a decision. I think it is unfortunate
when the good efforts of those who have worked to enhance a product are
compromised by the unsavory tactics of others seeking to promote it.
- nick
------------------------------
Date: Sun, 21 Aug 88 17:56:44
From: ZZZO%DHVRRZN1.BITNET@CUNYVM.CUNY.EDU
Subject: ELIZA
Date: 21 August 1988, 17:53:57 MEZ
From: Wolfgang Zocher (0511) 762-3684 ZZZO at DHVRRZN1
To: AILIST at AI.AI.MIT
Subject: Need for ELIZA
For the purpose of demonstration in a Lisp-course I need a Common Lisp
version of the ELIZA (Doctor) program (evtl. with Scripts)...
can anyone help me???
WZ (ZZZO at DHVRRZN1)
------------------------------
End of AIList Digest
********************
