CPSC 631 - Intelligent Agents and Multi-Agent Systems

Spring 2011

Professor: Dr. Thomas R. Ioerger
Office: 322C Bright Bldg.
Phone: 845-0161
email: ioerger@cs.tamu.edu
office hours: make appointment by email

Meeting: Tues/Thurs, 11:10-12:25, 126 Bright

Course Web Page: http://www.cs.tamu.edu/faculty/ioerger/cs631-spr11/index.html

Prerequisites: CPSC 625 (Artificial Intelligence) - or an equivalent AI course as an undergrad

Textbooks

The textbook used in the course (which will be liberally supplemented with papers from the literature) will be:

Other good textbooks on agents:

Grading

The course will be primarily project-based, and grades will be based on the quality of the projects implemented.

Goals of this Course

To learn about extended topics in AI, with a focus on intelligent agents. Specifically, we will cover the following major topics/perspectives:

  1. Agents: characteristics, environments, and architectures
  2. Multi-Agent Systems and Cooperative Behavior
  3. Adaptive Agents - Reinforcement Learning


Schedule:

Tues, Jan 18: Agent characteristics, environments, and applications
Thurs, Jan 20: Agent architectures read Ch 1-2 in Wooldridge, Ch. 2 in Russell and Norvig
Tues, Jan 25: knowledge-based agents, elevator policies, wumpus world in prolog
Thurs, Jan 27: modal logics (temporal, Metatem, beliefs, intentions)
Tues, Feb 1: MDPs read Sec. 1 & 2 of Szepesvari
Thurs, Feb 3: value and policy iteration (read Sec 3.1)
Tues, Feb 8: temporal-difference learning (see also Sec. 7.1-7.3 in Barto and Sutton)
Thurs, Feb 10: learning in continuous spaces, LSTD, LSPI (read Sec 3.2, also see Lagoudakis and Parr)
Tues, Feb 15: SARSA, Q-learning, actor-critic (Sec. 6.4-6.6 in Barto and Sutton)
Thurs, Feb 17: game theory (read Ch. 6 in Wooldridge)
Tues, Feb 22: auctions; MDP Project due (read Sec 7.1-7.2 in Wooldridge)
Thurs, Feb 24: negotiation (read Sec 7.3 in Wooldrige); Zlotkin and Rosenschein (1989). Negotiation and Task Sharing Among Autonomous Agents in Cooperative Domains. IJCAI. (FIGURES)
Tues, Mar 1: class cancelled
Thurs, Mar 3: voting E. Ephrati and J.S. Rosenschein (1991). The clarke tax as a consensus mechanism among automated agents. AAAI, pp. 173-178.; also, first 10 pages of Ephrati and Rosenschein (1996). AI Journal.
Tues, Mar 8: coalition formation Zlotkin, G., and Rosenschein, J. S., "Coalition, Cryptography, And Stability: Mechanisms For Coalition Formation In Task Oriented Domains", Proceedings of the Twelfth National Conference on Artificial Intelligence, pp. 432-437, 1994.
Thurs, Mar 10: coalition formation O. Shehory and S. Kraus. Task allocation via coalition formation among autonomous agents. In Proceedings of IJCAI'95, pages 655-661, 1995.
(Spring Break, Mar 14-18)
Tues, Mar 22: Project #2 due; DCSPYokoo and Durfee (1998). The distributed constraint satisfaction problem: formalization and algorithms.
Thurs, Mar 24:computational marketsYgge and Akkermans (1996). Power load management as a computational market.
Tues, Mar 29price equilibria in markets Wellman, Walsh, Wurman, and Makie-Mason (1998). Some Economics of Market-Based Distributed Scheduling.
Thurs, Mar 31: combinatorial auctionsread in detail: Walsh, Wellman, and Ygge (2000). Combinatorial Auctions for Supply Chain Formation.
also skim this: Nisan (2000). Bidding and Allocation in Combinatorial Auctions.
Tues, Apr 5: contract netsSandholm (1993). An Implementation of the Contract Net Protocol based on Marginal Cost Calculations.
Thurs, Apr 7: leveled commitmentsSandholm and Lesser (2002).
Tues, Apr 12: swarms: ant-colony optimizationDorgio, Barattari and Schutze (2006). Ant Colony Optimization. Computational Intelligence.
Thurs, Apr 14: swarms: bees; particle-swarm optimization focus on this: Particle Swarm Optimization: Developments, Applications, and Resources (2001)
skim this: Lucic and Tedorovic (2003). COMPUTING WITH BEES: ATTACKING COMPLEX TRANSPORTATION ENGINEERING PROBLEMS. Intl Jour of Art Intell Tools.
Tues, Apr 19: swarms: flocking Craig Reynolds (1987). Flocks, Herds, and Schools: A Distributed Behavioral Model - boids
Spector et al. (2003). Emergence of Collective Behaviour. GECCO'03 - SwarmEvolve
Thurs, Apr 21: swarms: cellular automata overview
Chan et al. (2010) - read sections 1-2.1 (Game of Life)
Stephen Wolfram (1983). Cellular Automata
Tues, Apr 26: swarms: El Farol bar problem Whithead (2008). The El Farol Bar Problem Revisited: Reinforcement Learning in a Potential Game
Thurs, Apr 28 (last class): RoboCupStone and Veloso (1998). A Layered Approach to Learning Client Behaviors in the RoboCup Soccer Server
see also Peter Stone's papers around 2001
Thurs, May 4: final project due


References


Links

Elevator Links

MDP/RL Reading Material