tjhorton@utai.UUCP (12/01/87)
UNIVERSITY OF TORONTO
Cognitive Science Events, December 1987
------------------------------------------------------------------------------
December 1 (Tuesday)
Artificial Intelligence Seminar
David Poole (Computer Science, Waterloo)
"Defaults and Conjectures: Hypothetical Reasoning for Explanation
and Prediction"
Abstract: Classical logic has been criticised as a language for
common sense reasoning as it is monotonic. In this talk I wish to
argue that the problem is not with logic, but with how logic is
used. An alternate way to use logic is by using theory formation;
logic tells us what a theory implies, an inconsistency means that
the theory cannot be true of the world. I show how the simplest
form of theory formation, namely where the user supplies the
possible hypotheses, can be used as a basis for default reasoning
and model-based diagnosis. This is the basis of the "Theorist"
system being built at the University of Waterloo, which will be
discussed.
Galbraith Building, rm 244, 2pm
December 2 (Wednesday)
Ebbinghaus Empire (human cognition and memory)
David Sherry (Psychology)
"Memory and the hippocampus in food-storing birds"
Sidney Smith Hall, rm 570, 12noon
December 2 (Wednesday)
CACS luncheon colloquium ($3 for lunch)
Lynd Forguson (Philosophy)
"Common Sense and Meta-Cognition"
OISE building, 4th floor, rm 296, 12 noon
"CACS" is the Center for Applied Cognitive Studies at OISE
December 2 (Wednesday)
Zeno Swijtink (Buffalo)
"The Doings in Infering"
University College, rm 152, 4:10pm
December 3 (Thursday)
special Artificial Intelligence Seminar
Peter Cheeseman (RIACS: NASA Ames Research Center)
"Automatic Discovery of Classes"
Abstract: This talk describes a criterion, based on Bayes's
theorem, that defines the optimal set of classes (a
classification) for a given set of examples. This criterion does
not require that the number of classes be specified in advance;
this is determined by the data. This approach avoids the need for
the user to define an ad hoc ``similarity measure'' between
objects and allows both real-valued and category-valued attributes
to be used. It also makes clear that classes are discovered, not
invented as some authors have claimed. The Bayesian approach can
be extended to general model discovery in data, and so it is the
solution to a major part of the induction problem.
Sandford Fleming Building, rm 1101, 11:00am, Everyone is welcome.
December 4 (Friday)
Derek Besner (Waterloo Psychology)
"On othographies and pholologies: how to read the bold print"
Scarboro Campus, Council Chamber, 12:15pm
December 4 (Friday)
Linguistics seminar
Brendan S. Gillon (Toronto)
"A Parametric Approach to Word Order: Evidence from Sanskrit"
Robarts Library, room 6071, 3:30pm
December 7 (Monday)
McLuhan Program: Literacy Seminar Series
Mark Seidengerg (McGill)
"Word recognition: A new theory of acquisition, skill
performance, and dyslexia"
For more info, call 978-7026
McLuhan Coach House, 39A Queen's Park, 4pm
December 8 (Tuesday)
Edward Stabler (CIAR and Computer Science, Western)
[stabler@uwocsd.uwo.cdn]
"The Logic of Movement and Barriers"
Abstract: No natural language processing system has explicitly
represented and used any substantial part of recent Chomskian
syntax.
This talk will describe an attempt to remedy that situation.
We present a transparent and flexible logical formalization of
some principles of Chomsky's "Barriers" theory, viz. the
"Barriers" principles relevant to movement and bounding.
Since the proposed logical representation is as transparent as
possible, it preserves the modularity and notational simplicity of
the theory itself, and it defines exactly the class of grammatical
structures covered by the theory.
The movement relations are of particular interest because their
formalization is relatively difficult, and consequently they are
inadequately treated in most automated systems.
The flexibility of the proposed formalization is illustrated by
extension of the system to allow verb raising and the amalgamation
of verbs with their affixes.
Since first order proof techniques are relatively well understood,
our theory can be evaluated directly or transformed into forms
that are more feasibly managed by particular theorem provers,
where the transformations are provably sound and as complete as
the intended application requires.
Galbraith Building, rm 120, 3pm
December 9 (Wednesday)
Ebbinghaus Empire (human cognition and memory)
Gus Craik (Psychology)
"Effects of aging on working memory"
Sidney Smith Hall, rm 570, 12noon
December 9 (Wednesday)
CACS seminar
Richard Wolfe (MECA at OISE) and Peter Rowley (Comp Sci)
"Hypertext and categorization theory: applications for
qualitative data analysis"
This will be an open house sort of presentation on the advanced
computing project in the MECA department [ed: this may not be
cogsci stuff].
OISE building, 2nd floor, rm 213, 12noon
December 10 (Thursday)
Institute of Biomedical Engineering seminar
Michael Beddoes (Electrical Engineering, UBC)
"Some applications of Wigner-Ville transformations in Speech
Research"
Rosebrugh building, rm 412, 1pm
December 10 (Thursday)
Physiology seminar
Dr. Peter Ashby (Medicine and Playfair Neuroscience Unit)
"Can we deduce excitatory and inhibitory post-synaptic potentials
in human motor neurons"
Medical Sciences Building, rm 3227, 4pm
December 11 (Friday)
Erindale/McMaster Cognitive Seminars
Dr. Keith Horton
"The processing of spatially transformed text"
Erindale, South Building, rm 3129, 3pm
December 16 (Wednesday)
Ebbinghaus Empire (human cognition and memory)
Speaker and topic T.B.A.
Contact Colin MacLeod at 284-3218, or electronically at
macleod@lake.scar.toronto.edu for details.
January 11 (Monday)
McLuhan Literacy and Computation Seminar Series
Michel Paradis (Linguistics, McGill)
"Neurolinguistic Aspects of Japanese Reading"
McLuhan Coach House, 39A Queen's Park, 4pm
January 15 (Friday)
Linguistics seminar
Marie Therese-Vinet (Universite de Sherbrooke)
"Empty Pleonastics in Haitian Creole and a paramterized INFL"
(syntax within a generative framework)
Robarts Library, 6th floor, rm 6071, 3:30pm
January 19 (Tuesday)
Artificial Intelligence Seminar
Roger Browse (Queens)
Some topic in vision.
Sandford Fleming Building, rm 1105, 2pm
January 20 (Wednesday)
Psychology Colloquium
Peter Shizgal (Concordia)
"Neuromechanisms of reward in the rat: a psychophysical and
electrophysiological analysis"
Sidney Smith Hall, rm 2135, 4pm
The Scarborough Campus has begun a Neuroscience speakers program.
The talks are held on every other Tuesday at 1 pm in the Council
Chamber at Scarborough. Get on the Neuroscience Newsletter
mailing list to receive information about them (call 978-4894).
The Toronto Working Papers in Linguistics, Vol. 8, is now
available from the Linguistics Graduate Course Union, Department
of Linguistics, at $7/copy. The volume contains work by Barbara
Brunson (Constraints on Discontinuity), Karen A. Carlyle (Vowel
Sequence and Sonority in Breton), Lisa Lai Shen Cheng (On the
'Passive' Construction in Mandarin), B. Elan Dresher (In Defence
of UG), and Keren Rice (Inflection Inside Derivation: An
Athapaskan Case).
The "University Lectures in Vision 1988" series is now officially
scheduled for the week of May 16-19, 1988. Contact Professor
Peter Hallett (Physiology), at 978-4339, for information. Monday,
May 16, will be an all day symposioum on "Colour and Contour" at
the George Ignatieff Theatre on the St. George Campus. Lectures
will continue at various times Tuesday and Wednesday, and F. de
Monasterio is a special invited speaker.
------------------------------------------------------------------------------
Excerpts from "Toronto Intelligence", Vol 1, Iss 4
(A Forum for Research and Study in Cognitive Science around UofT)
Paper subscriptions to Toronto Intelligence, event announcements, contact:
tjhorton@ai.toronto.edu
--
Timothy J Horton (416) 979-3109 tjhorton@ai.toronto.edu (CSnet,UUCP,Bitnet)
Dept of Computer Science tjhorton@ai.toronto (other Bitnet)
University of Toronto, tjhorton@ai.toronto.cdn (EAN X.400)
Toronto, Canada M5S 1A4 {seismo,watmath}!ai.toronto.edu!tjhorton