Cornell Cognitive Studies Symposium

STATISTICAL LEARNING ACROSS COGNITION

April 13-14, 2002

Cornell University

Organized by Morten Christiansen

mhc27@cornell.edu

Click here for the symposium program

Talks will be in Hollis E. Cornell Auditorium in Goldwin Smith Hall. The lunch and poster session will be held at the Big Red Barn. All events will take place on the Cornell campus.

 

Few things happen at random.  Rather, events in our environment tend to follow quasi-regular patterns.  Most organisms have evolved statistical learning mechanisms that allow them, at least in part, to detect and appropriately respond to such patterns.  For example, in sensory perception neighboring regions in the perceptual space are highly correlated; a property that in the visual domain is relevant for detecting edges in visual scenes.  In human language, picking up on the frequency with which syllables co-occur may be helpful in discovering word boundaries, just as learning about the statistics of word co-occurrences appear to be helpful in determining syntactic structure.

Because of its ubiquity across cognition, statistical learning has emerged as as an important and rapidly growing area of research within the cognitive sciences.  This symposium will bring together an interdisciplinary group of researchers from psychology, linguistics, computer science, neuroscience, and philosophy to take stock of current research in statistical learning and its future prospects.

 

 

Speakers:

 

Curt Burgess (University of California, Riverside)
Transforming Real-world Language into Semantic Representations

 

Nick Chater (University of Warwick, U.K.)
Simplicity as a Principle for Language Acquisition

 

Shimon Edelman (Cornell University)
Problems Arising in Unsupervised Learning of Structure

 

David Field (Cornell University)
Insights into Statistical Learning through an Understanding of the Statistical Structure of the Visual World

 

Rob Goldstone (Indiana University)
Creating Perceptual Representations that Recreate the World

 

Scott Johnson (Cornell University)
Visual Statistical Learning in Infancy

 

Lillian Lee (Cornell University)
Knowledge-lean Approaches to Statistical Natural Language Processing

 

David Lewkowicz (New York State Institute for Basic Research)
Perception of Multimodal Sequential Structure in Human Infants

 

Elissa Newport (University of Rochester)
Statistical Learning in Human Adults, Infants, and Nonhuman Primates: Constraints on Learning Adjacent and Nonadjacent Regularities

 

Michael Spivey (Cornell University)
Some Cross-modal Interactions Between Language and Vision

 

 

There will be a poster session Sunday afternoon at which graduate students (and faculty) can present work that relates to statistical learning.  Each student poster needs to be sponsored by a faculty member. Abstracts for the posters will be posted on the symposium web site and will be due April 9, 2002. The abstracts should be about 300 words long and must include one to two sentences defining statistical learning in relation to the work being presented. Please submit all abstracts (including name of sponsoring faculty) to Linda LeVan (cogst@cornell.edu). To go to the poster page, click here

 

This web site is continuously updated. Please check back for additional information and newly posted abstracts.