Zhong-Lin Lu, University of Southern California

Functional MRI has become one of the major tools in studying the brain mechanisms of human behavior. The detailed and enormous amount of data from fMRI experiments present both challenges and opportunities for quantitative analysis and modeling. This symposium will consist of a brief introduction to fMRI, followed by a series of talks on the state of the art data analysis/modeling techniques and applications.

Bob Dougherty, John S. George, James Haxby, Zhong-Lin Lu, John Serences, and Keith J. Worsley

Tom Griffiths, University of California, Berkely
Jay Myung, Ohio State University

Recent work in statistics has resulted in a number of sophisticated Monte Carlo methods that can be used for evaluating high-dimensional integrals and sampling from complex distributions, such as Markov chain Monte Carlo, importance sampling, and particle filtering. The symposium will consist of a tutorial on theoretical foundations of these methods, followed by a series of research talks exploring how the methods can be used in estimating the parameters of psychological models, designing experiments, and explaining different aspects of cognition.

Nando de Freitas

Tom Griffiths, Janne Kujala, Jay Myung, and Jeff Rouder

Neil Bearden, University of Arizona
Zygmunt Pizlo, Purdue University

The symposium covers a range of topics that traditionally were classified as representing different research areas, such as decision making, problem solving, operations research, and artificial intelligence. Recent progress in these areas, both empirical and theoretical, led to increased interdisciplinary activity, in which students of one area address questions and use tools from other areas. The symposium, which consists of two parts, will start with a keynote talk by William Cook, who will present an overview of combinatorial optimization problems. The two parts are organized around recent developments in the area of problem solving, and Markov decision processes.

William Cook

Speakers for part I (Problem Solving): Yll Haxhimusa, Ryan Murphy, Cole Smith, and Ulrike Stege
Speakers for part II (Markov Decision Processes): Neil Bearden, Ido Erev, Michael Lee, Brian Stankiewicz, and Josh Tenenbaum (tentative)

Mike Jones, Indiana University

The last decade has seen remarkable progress with techniques for learning lexical representations from statistical redundancies in language (e.g., Latent Semantic Analysis). This symposium will highlight some of the recent advances in computational models of language, including learning and representation of word meaning, compositional semantics, and grammar induction. In addition, the role of perceptual grounding and techniques to integrate multi-sensory data (above and beyond text) into statistical learning models will be discussed. With the emergence of many new models, we will focus on the ability of these techniques to explain core cognitive phenomena.

Tom Landauer

Simon Dennis, Mike Jones, Amy Perfors, Mark Steyvers, and Chen Yu