Copy Link
Add to Bookmark
Report
Neuron Digest Volume 04 Number 20
Neuron Digest Tuesday, 8 Nov 1988 Volume 4 : Issue 20
Today's Topics:
Looking for references on connectionism and Protein Structure
Re: Looking for references on connectionism and Protein Structure
Re: Looking for references on connectionism and Protein Structure
Looking for references to work on connectionism & databases
Need some intros to Neural Networks...
Re: Need some intros to Neural Networks...
Re: Need some intros to Neural Networks...
Neural networks, intelligent machines, and the AI wall: Jack Gelfand
Re: Neuron Digest V4 #18 (MUSIC and PDP)
PDP prog:s for Macintosh??
CA-simulator for Suns
CA-simulator
Send submissions, questions, address maintenance and requests for old issues to
"neuron-request@hplabs.hp.com" or "{any backbone,uunet}!hplabs!neuron-request"
------------------------------------------------------------
Subject: Looking for references on connectionism and Protein Structure
From: JRICE@cs.tcd.ie
Date: Thu, 27 Oct 88 10:23:00 +0000
I am interested in applying connectionism to the generation of protein
secondary and tertiary structure from the primary amino acid sequence.
Could you please send me any relevent references.
Thanks,
John.
------------------------------
Subject: Re: Looking for references on connectionism and Protein Structure
From: terry (Terry Sejnowski <terry>)
Date: Thu, 27 Oct 88 09:19:18 -0400
The best existing method for predicting the secondary structure of a
globular protein is a neural network:
Qian, N. and Sejnowski, T. J. (1988) Predicting the secondary structure of
globular proteins using neural network models. Journal of Molecular
Biology 202, 865-884.
Our results also indicate that only marginal improvements on our
performance will be possible with local methods.
Tertiary (3-D) structure is a much more difficult problem for which there
are no good methods.
Terry
------------------------------
Subject: Re: Looking for references on connectionism and Protein Structure
From: "Evan W. Steeg" <steeg@ai.toronto.edu>
Date: Thu, 27 Oct 88 11:13:54 -0400
In addition to the Qian & Sejnowski work, a potent illustration of the
capabilities of (even today's rather simplistic) neural nets, there is:
Bohr, N., Bohr, J., Brunak, S., Cotterill, R.M.J., Lautrup, B., Norskov,
L., Olsen, O.H., and Petersen, S.B. (of the Bohr Institute and Tech. Univ.
of Denmark), "Revealing Protein Structure by Neural Networks", presented as
a poster at the Fourth International Symposium on Biological and Artificial
Intelligence Systems, Trento, Italy 1988. Presumably, a paper will be
published shortly. They use a feed-forward net and back-propagation, like
Qian and Sejnowski, but use separate nets for each of 3 kinds of protein
secondary structure, rather than a single net, and there are other
methodological differences as well.
Others, myself included, are using neural net techniques which exploit
global (from the whole molecule), in addition to local interactions. This
should, as Dr. Sejnowski pointed out, lead to more accurate structure
prediction. Results of this work will begin to appear within a couple of
months.
-- Evan
Evan W. Steeg (416) 978-7321 steeg@ai.toronto.edu (CSnet,UUCP,Bitnet)
Dept of Computer Science steeg@ai.utoronto (other Bitnet)
University of Toronto, steeg@ai.toronto.cdn (EAN X.400)
Toronto, Canada M5S 1A4 {seismo,watmath}!ai.toronto.edu!steeg
------------------------------
Subject: Looking for references to work on connectionism & databases
From: pratt@paul.rutgers.edu (Lorien Y. Pratt)
Date: Wed, 26 Oct 88 12:27:42 -0400
I am interested in work which might relate neural networks to databases
in any way. Please send me any relevant references.
Thanks,
Lori
------------------------------
Subject: Need some intros to Neural Networks...
From: spector@vx2.NYU.EDU (David HM Spector)
Organization: New York University
Date: 30 Oct 88 01:28:00 +0000
Can anyone point me to some good intro texts on neural networks?
Preferable some thing with some good examples in C...
Thanks,
_DHMS
David HM Spector New York University
Senior Systems Programmer Stern School of Business
ARPAnet: SPECTOR@GBA.NYU.EDU Academic Computing Center
USEnet:...!{allegra,rocky,harvard}!cmcl2!spector 90 Trinity Place, Rm C-4
HamRadio: N2BCA MCIMail: DSpector New York, New York 10006
AppleLink: D1161 CompuServe: 71260,1410 (212) 285-6080
"Capital punishment is our society's recognition of the sanctity of human life"
- Senator Orrin Hatch
------------------------------
Subject: Re: Need some intros to Neural Networks...
From: wizard@jclyde.cactus.org (John Onorato)
Organization: Capitalist Warmongers, Inc., Austin, TX
Date: 30 Oct 88 17:31:27 +0000
I, too need some good introductory material for AI. I am intrigued by the
idea of ai, and I understand neural networks are a part of ai. However, I
do not know of any good introductory books on the subject... could someone
please point me in the right direction? As in the previous post on this
subject, I would like to see some examples in C, if this is possible.
thanks.......
wizard
------------------------------
Subject: Re: Need some intros to Neural Networks...
From: demers@beowulf.ucsd.edu (David E Demers)
Organization: EE/CS Dept. U.C. San Diego
Date: 30 Oct 88 19:20:47 +0000
In article <1790001@vx2.NYU.EDU> spector@vx2.NYU.EDU (David HM Spector) writes:
>Can anyone point me to some good intro texts on neural networks?
>Preferable some thing with some good examples in C...
I'm not sure that I'd call it an intro text, but you will no doubt have
many people refer you to Parallel Distributed Processing, by Rumelhart,
McClelland and the PDP Research Group, MIT Press, 1986. Volume 3 is
entitled Explorations in Parallel Distributed Processing and comes with IBM
PC software and lots of examples. You can play with the code as much as
you like.
There are several introductory texts rumored; Addison-Wesley will be
publishing Hecht-Nielsen's book, Neurocomputing, "soon". This is
engineering oriented, but what I've seen doesn't involve any code for
simulating nets. It talks more about implementing neural nets in hardware
(and not just HNC hardware...). There will no doubt be many other books
soon, since this is now a hot topic. I hope (in vain, for certain), we
don't see the glut of garbage books like the PC industry gave us.
Dave DeMers
UCSD Dept of Computer Science and Engineering
La Jolla, CA 92093
(619) 534-6254
demers@cs.ucsd.edu
------------------------------
Subject: Neural networks, intelligent machines, and the AI wall: Jack Gelfand
From: pratt@paul.rutgers.edu (Lorien Y. Pratt)
Date: Wed, 26 Oct 88 09:15:09 -0400
This looks like it'll be an especially interesting and controversial talk
for our department. I hope you all can make it!
--Lori
Fall, 1988
Neural Networks Colloquium Series
at Rutgers
NEURAL NETWORKS, INTELLIGENT MACHINES AND THE AI WALL
Jack Gelfand
The David Sarnoff Research Center
SRI International
Princeton,N.J.
Room 705 Hill center, Busch Campus
Friday November 4, 1988 at 11:10 am
Refreshments served before the talk
When we look back at the last 25 years of AI research, we find
that there have been many new techniques which have promised to
produce intelligent machines for real world applications. Though
the performance of some of these machines is quite extraordinary,
very few have approached the performance of human beings for even
the most rudimentary tasks. We believe that this is due to the fact
that these methods have been largely monolithic, whereas biological
systems approach these problems by combining many different modes
of processing into integrated systems. A number of real and artificial
neural network systems will be discussed in terms of how knowledge
is represented, combined and processed in order to solve complex
problems.
------------------------------
Subject: Re: Neuron Digest V4 #18
From: eliot@phoenix.Princeton.EDU (Eliot Handelman)
Organization: Princeton University, NJ
Date: 01 Nov 88 09:21:57 +0000
>Subject: Music and PDP (II)
>>From: MUSICO%BGERUG51.BITNET@CUNYVM.CUNY.EDU
>Date: Wed, 19 Oct 88 15:23:00 +0100
>
>There are other papers on Music and PDP. See :
> J. Bharucha : "Neural Net Modeling of Music",
> M. Leman : "Sequential (Musical) Information Processing
> with PDP-Networks",
> B. Vercoe : "Hearing Polyphonic Music with the Connection Machine"
>in : Proceedings of the first workshop on Music and AI.
>
>Work is also done by C. Lischka (See the Proceedings of the ICMC 87 and the
>Arbeidspapiere der GMD). I have two other related papers ("Neural Net-
>works in Music Research" and "Massive Parallel Computer Methods in Music
>Research"). I currently revise these papers and they will very soon be
>available.
>
>I guess that many other people are working in the same direction.
>
>Marc Leman
>University of Ghent
>Institute for Psychoacoustics and Electronic Music
>Blandijnberg 2
>B-9000 GHENT
>Belgium
I happened to have recently studied the three papers that Mr Leman
mentions. I have some remarks to make about both Mr Leman's and Mr
Bharucha's papers.
To begin with, Mr. Leman notes, in "Sequential (Musical) Information
Processing with PDP-Networks," that there has been some recent interest in
the "possibility of processing sequential information." and promises to
show us how to "store and process" "musical concepts," which he calls
"constraint examples of time-varying information." Fair enough.
What is unclear to me is the nature of Mr Leman's research. Is he proposing
to adapt other research to his specific needs, or is he merely providing us
with a summary of extant research? Mr Leman writes, for example, that the
"sequential autoassociator [..] is similar to the class of pattern
associators described by [...] Kosko (1988) BUT [my emphasis] applied to
the temporal domain." Is the suggestion not clear that this application is
original? I was curious to see what sources Mr Leman was drawing upon, and
so I looked up some of his references. One of these was a paper by Bart
Kosko in the IEEE Transactions on Systems, Man and Cybernetics, Vol. 18 no.
1, Jan/Feb 1988, "Bidirectional Associative Memories," which appeared in
print just a few months before Mr. Leman's paper.
Kosko: "The natural suggestion then is to *memorize* the association
(Ai,Bi) by forming the correlation matrix or vector outer product AT/iBi.
[...] The next suggestion is to *superimpose* the m associations by simply
adding up the correlation matrix pointwise: [a formula is given]."
[Kosko,52]
Leman: "Associations can be memorized by taking the vector outer product of
the associated patterns and then superimpose them so that: [the same
formula is given]."
It seems quite clear that Mr Leman has read Mr Kosko's paper. Did he not
notice that Mr Kosko is specifically interested in applying his asscociator
to the temporal domain? In his abstract, Mr. Kosko writes: "Temporal
patterns are represented as ordered lists of binary/bipolar vectors and
stored in a temporal associative memory (TAM) ..." [Kosko, pg 49]. Just in
case there was any doubt that Mr Kosko is interested in music, he writes
that "...a sequence of binary vectors can represent a harmonized melody."
[pg 49]
Not fair, Mr. Leman. Credit where credit is due.
Mr Bharucha's paper is another matter. Mr Leman at least gave us a few
formulae; Mr Bharucha is content to inform us of the power of his various
architectures and is apparently unwilling to let us judge for ourselves.
He writes, for example:
"This simple architecture has extraordinary predictive power for harmonic
expectancies in Western music. It predicts the build up of a tonal context
over the course of a sequence and it preserves the functional ambiguity
of individual events, thereby supporting smooth modulations."
These claims are nothing less than extraordinary. What exactly is meant by
"Western music"? Would this include the Trauerzug of Act II of Parsifal,
for example? How does it predict the build up of a tonal context? How does
it "preserve the functional ambiguity of individual events"? In what way
does it "support smooth modulations"? Mr Bharucha feels that it is
unnecessary to substantiate any of these claims with a few examples; I feel
that it is.
Is Mr Bharucha providing us with new architectures? This seems doubtful.
In "Storage and Processing of Information in Distributed Associative Memory
Systems" by Kohonen et al., in "parallel Models of Associative Memory,"
Hinton & Anderson, eds. LEA, 1981, page 121, an architecture is discussed
that recalls temporal sequences with additional context inputs using
delayed input that closely resembles those proposed by Mr. Bharucha. If Mr.
Bharucha is claiming to do original research into the architecture of
temporal pattern associators, he is at least 7 years out of date; if he is
doing some sort of music research, he might credit us with some curiosity
as to the basis of his claims.
Eliot Handelman
Department of Music
Princeton University
Princeton, NJ 08540
------------------------------
Subject: PDP prog:s for Macintosh??
From: d83_sven_a@tekn01.chalmers.se (Sven (Sciz) Axelsson)
Organization: Chalmers Univ. of Technology, Gothenburg, Sweden
Date: 03 Nov 88 19:42:19 +0000
Does anyone know of a port of Rumelhart & McClelland's PDP programs to the
Macintosh? I am considering doing this myself, but if it's been done
already...
Sven Axelsson d83_sven_a@tekn01.chalmers.se
Dep:t of Linguistics
Univ. of Gothenburg
SWEDEN
------------------------------
Subject: CA-simulator for Suns
From: cgl%cardinal@lanl.gov (Chris Langton)
Date: Sun, 30 Oct 88 03:17:37 -0700
[[ EDITOR'S NOTE: I culled this from the CA mailing list (Cellular
Automata). I thought redaers of Neuron Digest might be intersted in a
close relative. The message following this provides a European distribution
point. -PM ]]
The Cellsim cellular automaton simulator version 1.0 for Sun workstations
is now available and can be obtained free of charge via anonymous ftp.
It will run on COLOR Sun-3 and Sun-4 workstations.
This is a window-based tool for experimenting with 1-D or 2-D cellular
automata, allowing interactive editing of both the cellular arrays and
the transition functions, and supports real-time analysis of data generated
during CA runs.
Version 1.0 is quite a substantial improvement over the beta version that
we distributed for testing early in the summer, and incorportates many of
the suggestions made by those who obtained and ran the beta version. Thanks
to everybody who sent suggestions - I am sure you will find the 1.0 version
a vast improvement over the beta version.
Here is a scriptfile that details the procedure for obtaining the simulator:
- ----------------------------------------------------------------------------
raven[1]mkdir cellsim
raven[2]cd cellsim
raven[3]ftp 128.165.96.120
Connected to 128.165.96.120.
220 cardinal FTP server (Version 4.118 Thu Dec 31 21:21:26 MST 1987) ready.
Name (128.165.96.120:cgl): anonymous
331 Guest login ok, send ident as password.
Password: <TYPE SEVERAL RANDOM CHARACTERS HERE - NOT JUST RETURN!>
230 Guest login ok, access restrictions apply.
ftp> cd pub
250 CWD command successful.
ftp> ls
200 PORT command successful.
150 Opening data connection for /bin/ls (128.165.96.124,1785) (0 bytes).
Imagetool.tar.Z
cellsim_1.0.tar
comp.protocols.appletalk
lwsrv-problem
mac
ps.tar.Z
termcap.iris
ypserv
226 Transfer complete.
112 bytes received in .17 seconds (.64 Kbytes/s)
ftp> binary
200 Type set to I.
ftp> get cellsim_1.0.tar
200 PORT command successful.
150 Opening data connection for cellsim_1.0.tar (128.165.96.124,1786) (1916928 bytes).
226 Transfer complete.
local: cellsim_1.0.tar remote: cellsim_1.0.tar
1916928 bytes received in 17 seconds (1.1e+02 Kbytes/s)
ftp> bye
221 Goodbye.
raven[4]ls -l
total 1880
- -rw-r--r-- 1 cgl 1916928 Oct 30 02:32 cellsim_1.0.tar
raven[5]tar xvf cellsim_1.0.tar
x ./V1.0/mod8.v8, 32768 bytes, 64 tape blocks
x ./V1.0/cell.sun3, 884736 bytes, 1728 tape blocks
x ./V1.0/cell.sun4, 73728 bytes, 144 tape blocks
x ./V1.0/colors.64x, 4096 bytes, 8 tape blocks
x ./V1.0/src/cell.c, 17652 bytes, 35 tape blocks
x ./V1.0/src/cell.def, 1218 bytes, 3 tape blocks
x ./V1.0/src/cell.icon, 1933 bytes, 4 tape blocks
x ./V1.0/src/cell.param, 1472 bytes, 3 tape blocks
x ./V1.0/src/celldata.c, 7865 bytes, 16 tape blocks
x ./V1.0/src/cellmoore.c, 8164 bytes, 16 tape blocks
x ./V1.0/src/cellscr.c, 26415 bytes, 52 tape blocks
x ./V1.0/src/cellsock.c, 2056 bytes, 5 tape blocks
x ./V1.0/src/cellvonn.c, 8509 bytes, 17 tape blocks
x ./V1.0/src/celllin.c, 6480 bytes, 13 tape blocks
x ./V1.0/src/makefile, 662 bytes, 2 tape blocks
x ./V1.0/src/source.doc, 2452 bytes, 5 tape blocks
x ./V1.0/src/README, 2446 bytes, 5 tape blocks
x ./V1.0/mod5.v8, 32768 bytes, 64 tape blocks
x ./V1.0/codd.v8, 32768 bytes, 64 tape blocks
x ./V1.0/cross.64x, 4096 bytes, 8 tape blocks
x ./V1.0/README, 1039 bytes, 3 tape blocks
x ./V1.0/cell.doc, 22917 bytes, 45 tape blocks
x ./V1.0/rainbow.cmap, 64 bytes, 1 tape blocks
x ./V1.0/mod3.v8, 32768 bytes, 64 tape blocks
x ./V1.0/coddemit.64x, 4096 bytes, 8 tape blocks
x ./V1.0/arch.doc, 2362 bytes, 5 tape blocks
x ./V1.0/ckrbrd.64x, 4096 bytes, 8 tape blocks
x ./V1.0/life.m2, 512 bytes, 1 tape blocks
x ./V1.0/mod8.l8, 512 bytes, 1 tape blocks
x ./V1.0/mktrans.c, 2648 bytes, 6 tape blocks
x ./V1.0/socket/datagraph.c, 6152 bytes, 13 tape blocks
x ./V1.0/socket/dataread.c, 3392 bytes, 7 tape blocks
x ./V1.0/socket/README, 474 bytes, 1 tape blocks
x ./V1.0/nodes.64x, 4096 bytes, 8 tape blocks
x ./V1.0/firing.l8, 512 bytes, 1 tape blocks
x ./V1.0/ggun.64x, 4096 bytes, 8 tape blocks
x ./V1.0/learn2.64x, 4096 bytes, 8 tape blocks
x ./V1.0/loop.64x, 4096 bytes, 8 tape blocks
x ./V1.0/life.m4, 262144 bytes, 512 tape blocks
x ./V1.0/tunnel.v8, 32768 bytes, 64 tape blocks
x ./V1.0/heart.v8, 32768 bytes, 64 tape blocks
x ./V1.0/.indent.pro, 28 bytes, 1 tape blocks
x ./V1.0/loop.v8, 32768 bytes, 64 tape blocks
x ./V1.0/brain.m4, 262144 bytes, 512 tape blocks
x ./V1.0/learn1.64x, 4096 bytes, 8 tape blocks
x ./V1.0/firing.128, 128 bytes, 1 tape blocks
x ./V1.0/firing.64, 64 bytes, 1 tape blocks
x ./V1.0/firing.256, 256 bytes, 1 tape blocks
x ./V1.0/block.64x, 4096 bytes, 8 tape blocks
raven[6]ls -l V1.0
total 1776
- -rw-r--r-- 1 cgl 1039 Sep 2 17:34 README
- -rw-r--r-- 1 cgl 2362 Sep 30 18:16 arch.doc
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 block.64x
- -rw-r--r-- 1 cgl 262144 Sep 2 09:13 brain.m4
- -rw-r--r-- 1 cgl 22917 Oct 30 01:58 cell.doc
- -rwxr-xr-x 1 cgl 884736 Sep 2 09:13 cell.sun3*
- -rwxr-xr-x 1 cgl 73728 Sep 2 09:13 cell.sun4*
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 ckrbrd.64x
- -rw-r--r-- 1 cgl 32768 Sep 2 09:33 codd.v8
- -rw-r--r-- 1 cgl 4096 Sep 2 09:33 coddemit.64x
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 colors.64x
- -rw-r--r-- 1 cgl 4096 Sep 2 09:36 cross.64x
- -rw-r--r-- 1 cgl 128 Sep 2 09:13 firing.128
- -rw-r--r-- 1 cgl 256 Sep 2 09:13 firing.256
- -rw-r--r-- 1 cgl 64 Sep 2 09:13 firing.64
- -rw-r--r-- 1 cgl 512 Sep 2 09:13 firing.l8
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 ggun.64x
- -rw-r--r-- 1 cgl 32768 Sep 2 09:13 heart.v8
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 learn1.64x
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 learn2.64x
- -rw-r--r-- 1 cgl 512 Sep 2 09:13 life.m2
- -rw-r--r-- 1 cgl 262144 Sep 2 09:13 life.m4
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 loop.64x
- -rw-r--r-- 1 cgl 32768 Sep 2 09:13 loop.v8
- -rw-r--r-- 1 cgl 2648 Sep 2 14:55 mktrans.c
- -rw-r--r-- 1 cgl 32768 Sep 2 09:33 mod3.v8
- -rw-r--r-- 1 cgl 32768 Sep 2 09:33 mod5.v8
- -rw-r--r-- 1 cgl 512 Sep 2 09:13 mod8.l8
- -rw-r--r-- 1 cgl 32768 Sep 2 09:13 mod8.v8
- -rw-r--r-- 1 cgl 4096 Sep 2 09:13 nodes.64x
- -rw-r--r-- 1 cgl 64 Sep 2 09:13 rainbow.cmap
drwxr-xr-x 2 cgl 512 Sep 2 15:47 socket/
drwxr-xr-x 2 cgl 512 Oct 30 02:03 src/
- -rw-r--r-- 1 cgl 32768 Sep 2 09:13 tunnel.v8
raven[7]rm cellsim_1.0.tar
rm: remove cellsim_1.0.tar? y
- -----------------------------------------------------------------------------
After you have obtained the tar file and extracted the files, read the cell.doc
file for instructions on how to use the simulator.
Notice that we are including the source code with this release. This is in the
hope that others will add on features that they find useful.
Please address all comments, suggestions, or queries to me at the address below.
Chris Langton
Center for Nonlinear Studies Phone: 505-665-0059
MS B258 Email: cgl@LANL.GOV
Los Alamos National Laboratory
Los Alamos, New Mexico
87545
------------------------------
Subject: CA-simulator
From: T F Buckley <buckley%dcs.leeds.ac.uk@NSS.Cs.Ucl.AC.UK>
Date: Tue, 08 Nov 88 14:55:43 +0000
I have arranged with Chris Langton to act as a distribution point for his
Cellsim cellular automation simulator, (see his message to CA on Oct 30.)
This is for those in Europe who cannot ftp across to the USA. Chris is
sending me a tape with the files, in the meantime, those of you in Europe
who are interested in getting a copy let me know. My bitnet/earn address
is
buckley%uk.ac.leeds.dcs@UKACRL
If this doesn't work use CA to make the initial contact, hopefully this
will not place a strain on the network.
Tom Buckley
------------------------------
End of Neurons Digest
*********************
