Seminar on Computational Learning and Adaptation


The Problem

Over the past decade, research on computational approaches to learning and adaptation has emerged as a central topic in many disciplines, including artificial intelligence, molecular biology, cognitive psychology, complexity theory, decision theory, pattern recognition, and statistics. Unfortunately, researchers in these paradigms do not communicate as often as they might, leading to duplicated effort and missed insights that can come from interdisciplinary exchange.

The Response

The Seminar on Computational Learning and Adaptation is designed to improve communication among the local researchers with interests in computational approaches to learning and adaption, broadly defined. Talks cover a variety of methods - case-based learning, decision-tree induction, genetic algorithms, neural networks, and probabilistic algorithms - and take different approaches to evaluation - applied, experimental, theoretical, and psychological. Open discussion aims to establish a common language and increase the chances of future collaborations.

Logistics

During the Fall quarter of 2001-2002, the seminar will usually meet in Cordura 100 on thursdays from 4:15PM to 5:30PM. Cordura Hall is one of CSLI's (Center for the Study of Language and Information) buildings on the corner of Campus Drive and Panama Street (map). To reach Cordura 100, enter through the building's main doors, which are opposite Campus Drive and adjacent to Ventura Hall. Turn right into a short hall that ends in the meeting room.
 

Schedule for Fall Quarter 2002

Date Topic Speaker
October 4, 2001 Aprocedureforunsupervisedlexiconlearning Anand Venkataraman
Speech Technology and Research Laboratory, SRI International, Menlo Park, CA
October 11, 2001 Knowledge Acquisition using Relational Learning Ichise Ryutaro
CSLI, and National Institute of Informatics
October 18, 2001 On-Line Learning of Undirected Sparse n-grams Karl Pfleger
Computer Science Department, Stanford University
October 25, 2001 Sources of Success for Information Extraction Methods Joseph Smarr
Symbolic Systems Program, Stanford University
November 12, 2001 NeuroGrid: Learning to Search the Internet Sam Joseph
NeuroGrid Consulting
November 29, 2001 Learning Structure from Sequences, with Applications in a Digital Library Ian H. Witten
Department of Computer Science, University of Waikato, Hamilton, New Zealand
December 6, 2001 Analysing Stochastic Language Models Rune Lyngso
Baskin School of Engineering, University of California at Santa Cruz

 

Please pass on this information to other local researchers who might be interested in participating. If you would like to be added to the seminar mailing list, or if you are interested in giving a talk in the seminar, send email to dgs@stanford.edu.

Past Schedules

Return to Computational Learning Laboratory home page.