Graduate Summer School: Probabilistic Models of Cognition: The Mathematics of Mind

July 9 - 26, 2007

Schedule

All times in this Schedule are Pacific Time (PT)

Monday, July 9, 2007

Morning Session
8:00 - 8:45 Check-in & Continental Breakfast
8:45 - 9:00 Introductory Remarks
9:00 - 10:00
Josh Tenenbaum (Massachusetts Institute of Technology)

Introduction to Probabilistic Models of Cognition

10:00 - 10:15 Break
10:15 - 11:15
Josh Tenenbaum (Massachusetts Institute of Technology)

Basic Bayes: model fitting, model selection, and model averaging

11:15 - 11:30 Break
11:30 - 12:30
12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Introduction to Graphical Models

3:00 - 4:00
Robert Jacobs (University of Rochester)

Bayesian Decision Theory

4:00 - 4:30 Break
4:30 - 5:30
Ying Nian Wu (UCLA)

Information Theory

5:30 - 7:00 Opening Reception hosted by IPAM (IPAM Building)

*Optional Evening Session - Nick Chater, Tom Griffiths, Josh Tenenbaum, "Subjective probability and Bayesian foundations"

Matlab tutorial, by Amy Perfors

Tuesday, July 10, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Alan Yuille (University of California, Los Angeles (UCLA))

Graphical Models in Vision

10:00 - 10:15 Break
10:15 - 11:15
Alan Yuille (University of California, Los Angeles (UCLA))

HMMs and inference in graphical models I

11:15 - 11:30 Break
11:30 - 12:30
Robert Jacobs (University of Rochester)

Ideal Observers and Ideal Actors

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45
3:00 - 4:00
Alan Yuille (University of California, Los Angeles (UCLA))

HMMs and inference in graphical models II

4:00 - 4:30 Break
4:30 - 5:30
Robert Jacobs (University of Rochester)

Sensory Cue Combination

*Optional Evening Session - Introduction to software for Bayesian modeling: SamIAm, Bayes Net Toolbox (BNT)

Wednesday, July 11, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Zoubin Ghahramani (University of Cambridge)

Probabilistic generative models and unsupervised learning I

10:00 - 10:15 Break
10:15 - 11:15
Zoubin Ghahramani (University of Cambridge)

Probabilistic generative models and unsupervised learning II

11:15 - 11:30 Break
11:30 - 12:30
Zoubin Ghahramani (University of Cambridge)

Nonparametric Bayesian Models

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Zoubin Ghahramani
Questions on generative models and nonparametric Bayes
3:00 - 4:00
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Modern Monte Carlo methods I: Importance sampling and sequential Monte Carlo

4:00 - 4:30 Break
4:30 - 5:30
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Modern Monte Carlo methods II: Markov chain Monte Carlo

*Optional Evening Session - Nick Chater, "Scale invariance in human cognition"
PowerPoint Presentation

Thursday, July 12, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:30
Zoubin Ghahramani (University of Cambridge)

Bayesian supervised learning and semi-supervised learning

10:30 - 11:00 Break
11:00 - 12:30
12:30 - 1:30 Lunch (on your own)

Charles Kemp, Amy Perfors, Josh Tenenbaum
Learning overhypotheses with hierarchical Bayesian models

Afternoon Session
1:45 - 2:45 Breakout Sessions

Zoubin Ghahramani, Yingnian Wu
Technical questions on Bayesian models

Tom Griffiths
MCMC with humans, and other questions on MCMC

"Practical experiments with simple Bayesian models", Amy Perfors
3:00 - 4:15
Josh Tenenbaum (Massachusetts Institute of Technology)

Hierarchical Bayesian models of human inductive learning

4:15 - 4:30 Break
4:30 - 5:30

*Optional Evening Session - Michael Lee, "Bayesian model theoretics and BUGS"

Friday, July 13, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 9:45
Ying Nian Wu (UCLA)

Causal Models

9:50 - 10:45
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Graphical models and human causal learning

10:45 - 11:15 Break
11:15 - 12:00
12:05 - 12:50
Patricia Cheng (University of California, Los Angeles (UCLA))

Causal Reasoning in Humans and Rats

12:50 - 2:00 Lunch (on your own)

Breakout Sessions

Ed Vul
"Bayes Net Toolbox causal model demo"

Afternoon Session
2:00 - 3:00
Alison Gopnik (University of California, Berkeley (UC Berkeley))

Causal Learning in Children

3:15 - 4:00
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Modeling causal learning in children

4:00 - 4:30 Break
4:30 - 5:30
Keith Holyoak (University of California, Los Angeles (UCLA))

From Causal Models to Analogical Inference


Monday, July 16, 2007

Morning Session
8:00 - 9:00 Check-in & Continental Breakfast
9:00 - 10:00
Stuart Russell (University of California, Berkeley (UC Berkeley))

Probability for Worlds with Things in Them

10:00 - 10:15 Break
10:15 - 11:15
Brian Milch (Massachusetts Institute of Technology)

Relational Probability Models

11:15 - 11:30 Break
11:30 - 12:30
Brian Milch (Massachusetts Institute of Technology)

Unknown Objects and BLOG

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Sessions

Stuart Russell, Brian Milch
Questions on first-order probability and BLOG
3:00 - 4:00
Josh Tenenbaum (Massachusetts Institute of Technology)

First-order probabilistic models in human cognition

4:00 - 4:30 Break
4:30 - 5:30
Charles Kemp (Massachusetts Institute of Technology)

Learning relational theories

*Optional Evening Session - Stanley Kok, Markov Logic Networks (MLNs) and the Alchemy system
Presentation (PowerPoint)

Tuesday, July 17, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
10:00 - 10:15 Break
10:15 - 11:15
Stuart Geman (Brown University)

Hierarchy and Reusability in Image Analysis II

11:15 - 11:30 Break
11:30 - 12:30
Alan Yuille (University of California, Los Angeles (UCLA))

Image Parsing

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Stuart Geman/Alan Yuille
Questions on grammars for image parsing

Stanley Kok
Questions on MLNs and Alchemy
3:00 - 4:00
Fei-Fei Li (Princeton University)

Discovering Meaning in the Visual World

4:00 - 4:30 Break
4:30 - 5:30
Dan Kersten (University of Minnesota, Twin Cities)

Human Object Perception: Bottom-up and Top-down

*Optional Evening Session: Stuart Russell, Brian Milch, Josh Tenenbaum, Charles Kemp, et al: Discussion on connections between Cognitive Science and AI.

Wednesday, July 18, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Mark Johnson (Brown University)

An introduction to grammars and parsing for language 1

10:00 - 10:15 Break
10:15 - 11:15
Mark Johnson (Brown University)

An introduction to grammars and parsing for language 2
Updated Presentation

11:15 - 11:30 Break
11:30 - 12:30
Roger Levy (University of California, San Diego)

Probabilistic sentence processing 1

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Roger Levy, Mark Johnson
Questions on grammars for language, and connections between human and machine language research

Noah Goodman
Probabilistic grammars for concept learning
3:00 - 4:00
Roger Levy (University of California, San Diego)

Probabilistic sentence processing 2

4:00 - 4:30 Break
4:30 - 5:30
Mark Steyvers (University of California, Irvine (UCI))

Semantic Representations with Probabilistic Topic Models

5:30 - 7:00 Reception hosted by IPAM (IPAM Building)

Thursday, July 19, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Peter Dayan (University College London)

Models of conditioning

10:00 - 10:15 Break
10:15 - 11:15
John Kruschke (Indiana University)

Locally Bayesian Learning

11:15 - 11:30 Break
11:30 - 12:30
12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Stuart Russell, Peter Dayan
Questions and dialog on reinforcement learning
3:00 - 4:00
Stuart Russell (University of California, Berkeley (UC Berkeley))

Reinforcement Learning

4:00 - 4:30 Break
4:30 - 5:30

Friday, July 20, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
10:00 - 10:15 Break
10:15 - 11:15
Peter Dayan (University College London)

Neural representation of value, reward and expectation

11:15 - 11:30 Break
11:30 - 12:30
12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Peter Dayan, Josh Tenenbaum, Angela Yu, & others
Neural representations of probability distributions

Jeffrey Beck (University of Rochester)
Bayesian Inference, Action Selection, and Learning with Probabilistic Population Codes
Presentation (PDF File)
3:00 - 4:00
Stuart Russell (University of California, Berkeley (UC Berkeley))

Hierarchical reinforcement learning

4:00 - 4:30 Break
4:30 - 5:30

Monday, July 23, 2007

Morning Session
8:00 - 9:00 Check-in & Continental Breakfast
9:00 - 10:00
10:00 - 10:15 Break
10:15 - 11:15
11:15 - 11:30 Break
11:30 - 12:30
Konrad Koerding (Northwestern University Medical School)

Using Decision Theory to understand Motor Control

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 - 2:45 Breakout Session

Craig Fox (UCLA)
Prospect theory and support theory
Reading: "Belief and Preference in Decision Under Uncertainty", by Craig R. Fox & Kelly E. See
Presentation (PDF File)
3:00 - 4:00
Josh Tenenbaum (Massachusetts Institute of Technology)

Rational analysis of human memory and prediction

4:00 - 4:30 Break
4:30 - 5:30
Rich Shiffrin (Indiana University)

Bayesian models of memory retrieval


Tuesday, July 24, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Amy Perfors (Massachusetts Institute of Technology)

Grammar induction in language

10:00 - 10:15 Break
10:15 - 11:15
Alan Yuille (University of California, Los Angeles (UCLA))

Grammer induction in vision

11:15 - 11:30 Break
11:30 - 12:15
Josh Tenenbaum (Massachusetts Institute of Technology)

Cognitive development as a computational challenge: the grammar analogy

12:15 - 1:15 Lunch (on your own)

Afternoon Session
1:30 - 2:30 Breakout Session

Opportunities for mathematicians and statisticians in cognitive science and neuroscience, with Alan Yuille, Mark Green, Yingnian Wu, Noah Goodman, others

Faculty Center
2:45 - 3:30
Charles Kemp (Massachusetts Institute of Technology)

The development of structured representations

3:30 - 3:50 Break
3:50 - 4:25
Tom Griffiths (University of California, Berkeley (UC Berkeley))

Development of Causal Theories

4:25 - 4:45 Break
4:45 - 5:30
Noah Goodman (Massachusetts Institute of Technology)

Learning to be causal

7:30 Discussion and Q & A on grammar induction and unsupervised learning for structured statistical models with Alan Yuille, Josh Tenenbaum, Noah Goodman and Tom Griffiths

Math Sci 4000A

Wednesday, July 25, 2007

Morning Session
8:00 - 9:00 Continental Breakfast
9:00 - 10:00
Konrad Koerding (Northwestern University Medical School)

Causal Inference in Motor Learning and Adaptation

10:00 - 10:15 Break
10:15 - 11:15
11:15 - 11:30 Break
11:30 - 12:30
Jerome Busemeyer (Indiana University)

Challenges for Bayes

12:30 - 1:30 Lunch (on your own)

Afternoon Session
1:45 Symposium (Future Challenges): Josh Tenenbaum, Tom Griffiths, Alan Yuille, Keith Holyoak, Rich Shiffrin

Thursday, July 26, 2007

Morning Session
8:00 - 9:00 Continental Breakfast

Student Research Symposium