- Interdisciplinary Program in Cognitive Science, Seoul National University
- Instructor: Dr. Joon Shik Kim (Room: 302-314-1, E-mail: jskim@bi.snu.ac.kr, Tel: 880-1847, H.P.: 010-2838-0324)
- TA: Chung-Yeon Lee (Room: 302-314-1, E-mail: cylee@bi.snu.ac.kr, Tel: 880-1847)
- Classroom: 14-207-1
- Time: Friday 09:00-12:00
- Objectives
- Mathematical and statistical modeling of intelligence
- Study of an information processing system based on the masterpiece papers in early days
- Study of inference algorithms of Markov chain Monte Carlo (MCMC)
- Text book
- Unsupervised Learning: Foundations of Neural Computation, edited by Geoffrey Hinton and Terrence J. Sejnowski, The MIT Press, 1999
- Evaluation
- Paper presentation (20%)
- Research presentation I (15%)
- Research presentation II (15%)
- Report of a term project (30%)
- Participation in discussion (20%)
- Announcement
- Presentation will be evaluated by other students
- A term project
- The project consists of performing a cognitive science experiment and writing a final report
- Main reference for a project will be presented at the paper presentation class
- References
- Computing machinery and intelligence, AM Turing, Mind, 59(236): 433-460, 1950.
- Preliminary discussion of the logical design of an electronic computing instrument, AW Burks, HH Goldstine, and J von Neumann, Papers of John von Neumann on computing and computer theory: MIT press, 97-142, 1987.
- A theory for archicortex, D Marr, Philosophical transactions of the royal society of London, 262: 3-81, 1971.
- Theory of communication, CE Shannon, ACM mobile communications review, 5(1): 3-55, 2001.
- A global minimization algorithm based on a geodesic of a Lagrangian formulation of Newtonian dynamics, JS Kim, JC Kim, J O, and BT Zhang, Neural Processing Letters, 26(2): 121-131, 2007.
- Markov chain sampling methods for Drichlet process mixture models, RM Neal, Journal of computational and graphical statistics, 9(2): 249-265, 2000.
- Hierarchical Dirichlet processes, YW Teh, MI Jordan, MJ Beal, DM Blei, Journal of the American statistical association, 101(476): 1566-1581, 2005.
- Z Ghahramani, TL Griffiths, and P Sollich, Bayesian nonparametric latent feature model, proc. Valencia / ISBA 8th world meeting on Bayesian statistics, June 1st-6th, 2006.
- NL Roux and Y Bengio, Representational power of restricted Boltzmann machines and deep belief networks, Neural Computation, 20: 1631-1649, 2008.
- An evolutionary Monte Carlo algorithm for predicting DNA hybridization, JS Kim, JW Lee, YK Noh, JY Park, DY Lee, KA Yang, YG Chai, JC Kim, and BT Zhang, Biosystems, 91(1):69-75, 2008.
- Schedule
Date |
Paper |
Speaker |
Slides |
9.2 |
Paper written by Turing (Ref. [1]) |
Joon Shik Kim |
|
9.9 |
Paper written by Neumann (Ref. [2]) |
Joon Shik Kim |
|
9.16 |
Paper written by Marr (Ref. [3]) |
Joon Shik Kim |
|
9.23 |
Paper written by Shannon (Ref. [4]) |
Joon Shik Kim |
|
9.30 |
Geodesic of a Lagrangian (Ref. [4]) |
Joon Shik Kim |
|
10.7 |
Paper Presentation |
Students |
|
10.14 |
Markov chain sampling (Ref. [6]) |
Joon Shik Kim |
|
10.21 |
Hierachical Dirichlet process (Ref. [7]) |
Joon Shik Kim |
|
10.28 |
Indian buffet process (Ref. [8]) |
Joon Shik Kim |
|
11.4 |
Research Presentation I |
Students |
|
11.11 |
Restricted Bolzmann machine (Ref. [9]) |
Joon Shik Kim |
|
11.18 |
DNA computing (Ref. [10]) |
Joon Shik Kim |
|
11.25 |
Research presentation II |
Students |
|
12.2 |
Due day of the final report |
|
|
|