harnad@phoenix.Princeton.EDU (Stevan Harnad) (04/24/91)
Archive-name: ai/neural-nets/harnad-categorical-perception/1991-04-23
Archive-directory: princeton.edu:/pub/harnad/ [128.112.128.1]
Original-posting-by: harnad@phoenix.Princeton.EDU (Stevan Harnad)
Original-subject: Categorical Perception and Neural Nets
Reposted-by: emv@msen.com (Edward Vielmetti, MSEN)
The following two tech reports are available by anonymous ftp from
directory /pub/harnad on princeton.edu. Full ftp instructions follow
the abstracts.
-------------------------------------------------
(1) Categorical Perception and the Evolution
of Supervised Learning in Neural Nets
S Harnad*, SJ Hanson*,** & J Lubin*
*Princeton University
**Siemens Research Center
[Presented at 1991 AAAI Symposium on Symbol Grounding: Problem and
Practice]
ABSTRACT: Some of the features of animal and human categorical
perception (CP) for color, pitch and speech are exhibited by neural
net simulations of CP with one-dimensional inputs: When a backprop net
is trained to discriminate and then categorize a set of stimuli, the
second task is accomplished by "warping" the similarity space
(compressing within-category distances and expanding between-category
distances). This natural side-effect also occurs in humans and
animals. Such CP categories, consisting of named, bounded regions of
similarity space, may be the ground level out of which higher-order
categories are constructed; nets are one possible candidate for the
mechanism that learns the sensorimotor invariants that connect
arbitrary names (elementary symbols?) to the nonarbitrary shapes of
objects. This paper examines how and why such compression/expansion
effects occur in neural nets.
[Retrieve by anonymous ftp in binary mode as (compressed) file
harnad91.cpnets.Z from directory /pub/harnad on princeton.edu,
instructions below]
-----------------------------------------------------------------
(2) Connecting Object to Symbol in Modeling Cognition
Stevan Harnad
Department of Psychology
Princeton University
Princeton NJ 08544
[To appear in Clark, A. & Lutz, R. (Eds) (1992) "CONNECTIONISM IN
CONTEXT," Springer-Verlag]
Connectionism and computationalism are currently vying for hegemony in
cognitive modeling. At first glance the opposition seems incoherent,
because connectionism is itself computational, but the form of
computationalism that has been the prime candidate for encoding the
"language of thought" has been symbolic computationalism, whereas
connectionism is nonsymbolic. This paper examines what is and is not a
symbol system. A hybrid nonsymbolic/symbolic system will be sketched in
which the meanings of the symbols are grounded bottom-up in the
system's capacity to discriminate and identify the objects they refer
to. Neural nets are one possible mechanism for learning the invariants
in the analog sensory projection on which successful categorization is
based. "Categorical perception," in which similarity space is "warped"
in the service of categorization, turns out to be exhibited by both
people and nets, and may mediate the constraints exerted by the analog
world of objects on the formal world of symbols.
[Retrieve by anonymous ftp in binary mode as (compressed) file
harnad92.symbol.object.Z from directory /pub/harnad on princeton.edu]
To retrieve a file by ftp from a Unix/Internet site, type:
ftp princeton.edu
When you are asked for your login, type:
anonymous
For your password, type your full name
then change directories with:
cd pub/harnad
Then type:
binary
(This is for retrieving compressed files.)
To show the available files, type:
ls
Next, retrieve the file you want with (for example):
get filename.Z
When you have the file(s) you want, type:
quit
Next uncompress the file with:
uncompress filename.Z
Now the file will be called, simply,
filename
---
The above cannot be done from Bitnet directly, but there
is a fileserver called bitftp@pucc.bitnet that will do
it for you. Send it the one line message
help
for instructions (which will be similar to the above,
but will be in the form of a series of lines in an
email message that bitftp will then execute for you).
Stevan Harnad Department of Psychology Princeton University
harnad@clarity.princeton.edu / harnad@pucc.bitnet / srh@flash.bellcore.com
harnad@learning.siemens.com / harnad@elbereth.rutgers.edu / (609)-921-7771
-- comp.archives file verification
princeton.edu
total 2028
-rw-rw-rw- 1 116 13408 Apr 23 20:21 harnad91.cpnets.Z
-rw-rw-rw- 1 116 128568 Apr 23 15:45 bbs.dennett
-rw------- 1 116 187199 Apr 17 02:20 alist
-rw-rw-rw- 1 116 23075 Apr 10 15:35 harnad92.symbol.object.Z
-rw-r--r-- 1 116 108018 Mar 28 17:03 bbs.kenrick
-rw-rw-rw- 1 116 6071 Feb 5 02:24 psyc.arch.2.2
-rw-rw-rw- 1 116 18570 Feb 5 02:13 psyc.arch.2.1
-rw-rw-rw- 1 116 178531 Feb 2 20:43 bbs.thompson
-rw-rw-rw- 1 116 142600 Feb 2 20:43 bbs.greenfield
-rw-rw-rw- 1 116 131684 Feb 2 20:42 bbs.crain
-rw-rw-rw- 1 116 2803 Jan 24 18:39 psyc.background8.Z
-rw-r--r-- 1 116 4519 Jan 24 18:37 psyc.background5.Z
-rw-rw-rw- 1 116 4078 Jan 24 18:36 psyc.background2.Z
-rw-rw-rw- 1 116 1604 Jan 24 18:34 psyc.background1.Z
-rw-rw-rw- 1 116 66379 Jan 15 15:57 sg5.archive.Z
-rw-r--r-- 1 116 15 Jan 15 15:52 .forward
-rw-rw-rw- 1 116 10156 Jan 14 01:22 harnad90.dietrich.crit.Z
-rw-rw-rw- 1 116 8985 Jan 13 22:06 harnad90.skywriting.Z
-rw-r--r-- 1 116 70463 Jan 4 00:42 psyc.arch.1.16
-rw-rw-rw- 1 116 19419 Dec 3 22:23 psyc.arch.1.14
-rw-rw-rw- 1 116 37222 Dec 3 22:21 psyc.arch.1.15
-rw-rw-rw- 1 116 11807 Nov 28 04:08 harnad90.dyer.crit.Z
-rw-r--r-- 1 116 19235 Nov 8 03:54 harnad91.otherminds.Z
-rw-rw-rw- 1 116 15882 Nov 6 21:20 harnad82.neoconst.Z
-rw-rw-rw- 1 116 29006 Nov 6 21:20 harnad82.consciousness.Z
-rw-rw-rw- 1 116 26430 Oct 11 1990 psyc.arch.1.13
-rw-rw-rw- 1 116 13234 Sep 15 1990 psyc.arch.1.12
-rw-rw-rw- 1 116 40187 Aug 24 1990 psyc.arch.1.11
-rw-rw-rw- 1 116 24816 Aug 22 1990 harnad90.sgproblem.Z
-rw-rw-rw- 1 116 50151 Aug 9 1990 psyc.arch.1.10
-rw-rw-rw- 1 116 1974 Jul 24 1990 psyc.background7.Z
-rw-rw-rw- 1 116 8959 Jul 24 1990 psyc.background6.Z
-rw-rw-rw- 1 116 9831 Jul 24 1990 psyc.background4.Z
-rw-rw-rw- 1 116 2014 Jul 24 1990 psyc.background3.Z
-rw-rw-rw- 1 116 6732 Jul 20 1990 psyc.arch.1.9
-rw-rw-rw- 1 116 27866 Jul 20 1990 psyc.arch.1.8
-rw-rw-rw- 1 116 26239 Jul 20 1990 psyc.arch.1.1-7
-rw-rw-rw- 1 116 91689 Jul 20 1990 sg4.archive.Z
-rw-rw-rw- 1 116 116469 Jul 20 1990 sg3.archive.Z
-rw-rw-rw- 1 116 52046 Jul 20 1990 sg2.archive.Z
-rw-rw-rw- 1 116 132345 Jul 20 1990 sg1.archive.Z
-rw-rw-rw- 1 116 35660 Jul 20 1990 harnad89.searle.Z
-rw-rw-rw- 1 116 54097 Jul 20 1990 harnad87.categorization.Z
found harnad-categorical-perception ok
princeton.edu:/pub/harnad/