SNU Biointelligence Lab

4541.676 Action-Perception-Learning Cycles (Fall 2012)

Graduate Course in Computer Science and Engineering, Cognitive Science, and Brain Science

Course Description

How can the brain learn so fast, flexibly, and robustly? What representational mechanisms and organizational principles does the brain use? How can we apply these principles to constructing intelligent cognitive machines that learn like humans? To address these questions, it is important to observe that the brain is embodied with sensors and actuators, and interacts with its environment in a continuous perception-action cycle. Living in a dynamic environment under uncertainty requires the brain to learn moment by moment in real time and incrementally in this continuous, rapid perception-action cycle. In this course we review recent experimental and theoretical work on perception-action cycles and neural coding principles in the brain. We also study mathematical tools developed in information theory, control theory, and Bayesian statistics. The goal is to develop computational models of sequential learning processes — action-perception-learning cycle machines — that enable rapid, continuous, and reliable action and decision-making in a changing environment.

References

  1. Bayesian Time Series Models — Barber, Cemgil & Chiappa — 2011
  2. Kalman Filtering and Neural Networks — Haykin — 2001
  3. Beyond the Kalman Filter — Ristic, Arulampalam & Gordon — 2004

Grading Policy

구분 비율
Two Open-book Exams 70%
Paper Presentations 20%
Participation & Discussion 10%

Course Schedule

Week Date Topics
1 9/4, 9/6 From machine learning to brain-like cognitive learning · Brain as a physical, thermodynamic computer · Perception-action cycles and Carnot cycles · Models of action-perception-learning cycles
2 9/11, 9/13 Gordon et al.: Toward an integrated approach to perception and action (2011) · Fuster: The prefrontal cortex — Time is of the essence, Neuron 30:319–333 (2001)
3 9/18, 9/20 Pickering & Garrod: An integrated theory of language production and comprehension, BBS (2012) · Clark: Whatever next? Predictive brains, situated agents, BBS (2012)
4 9/25, 9/27 Fry: Neural statics and dynamics, Neurocomputing 65–66:455–462 (2005) · Reviews and discussion
5 10/2, 10/4 Friston: The free-energy principle: a rough guide to the brain?, TiCS (2009) · Dayan & Daw: Connections between computational and neurobiological perspectives on decision making, CABN (2008)
6 10/9, 10/11 Zahedi et al.: Higher coordination with less control — Information maximization, Adaptive Behavior 18(3–4):338–355 (2010) · Tishby & Polani: Information theory of decisions and actions (2010)
7 10/16, 10/18 Reviews and discussion · Exam 1
8 10/23, 10/25 Bayesian inference and estimation · Barber, Cemgil & Chiappa: Inference and estimation in probabilistic time series models [1]
9 10/30, 11/1 Barber, Cemgil & Chiappa: Inference and estimation in probabilistic time series models [1] (continued)
10 11/6, 11/8 Haykin: Kalman filters [2] · Extensions and variations
11 11/13, 11/15 Pouget et al.: Inference and computation with population codes, Ann. Rev. Neuroscience 26:381–410 (2003) · Knill & Pouget: The Bayesian brain, TiNS 27(12):712–719 (2004)
12 11/20, 11/22 A tutorial on particle filters [3] · Bayesian evolutionary algorithms
13 11/27, 11/29 Reviews and discussion · Exam 2
14 12/4, 12/6 Osborne et al.: The neural basis for combinatorial coding in a cortical population, J. Neuroscience 28:13522–13531 (2008) · Bach & Dolan: Knowing how much you don’t know, Nature Rev. Neuroscience 13:572–586 (2012)
15 12/11, 12/13 Reviews and discussion · Poster Workshop (Room 302-308)