CSCI 543: Software Multiagent Systems

Spring 2008: Mondays and Wednesdays 2:00 p.m. – 3:20 p.m. at Kaprielian Hall 140 (KAP140)

 

====================================================================

Course notes are available from the USC bookstore

Papers will be updated with new publications from 2008 as they become available

 

Syllabus

 

 

    PART I: Fundamentals of agents and multiagent systems

 

  1. (January 14): Course intro, syllabus, what is an intelligent agent

 

  1. (January 16): Beliefs, desires, intentions (BDI), Satisficing, Begin BDI logics

 

  1. (January 23): BDI logics continued, BDI architectures, reactive plans

 

  1. (January 28): Decision theory I; Making simple decisions under uncertainty; risk; risk averseness, risk neutrality; Sequential decisions under uncertainty, Markov decision problems; Value iteration; basic introduction to POMDPs

Suggested reading: Kaebling, Littman, Cassandra: Planning and Acting in Partially Observable Stochastic Domains, AIJ 101, 1998

 

  1. (January 30, Feb 4): Game theory: Normal form and extensive form games, Prisoner’s dilemma, Chicken,  Dominance, iterative dominance, Nash equilibrium; Mixed strategy Nash equilibrium, Iterative Prisonners dilemma, Stackelberg games, Bayesian Games, Harsanyi transformation

 

  1. (Feb 6): Auctions: First price, second price, sequential auctions

 

  1. (Feb 11):  Distributed constraint reasoning, Distributed constraint satisfaction

 

 

PART II: Collaboration: Agent-agent and agent-human interactions

 

  1. (February 13): Distributed constraint optimization I: Complete algorithms

 

  1. (February 20): Distributed constraint optimization II: Incomplete algorithms and recent theoretical results

 

  1. (February 25):   Distributed POMDPs

 

  1. (February 27):  Teamwork I: What is teamwork, team logic

 

  1. (March 3): Teamwork II: Practically implementing teamwork beyond joint persistent goals: representing team plans and roles in an agent architecture, addressing practical communication costs, team monitoring and recovery from failures. Introduction to decision theoretic approaches to teamwork.

 

  1. (March 5): Quiz I

 

  1. (March 10): Distributed POMDPs, hybrid approaches to teamwork

 

  1. (March 12): Panel discussion: Where are we going with agent technologies: are we on the right track

 

  1. (March 24): Team formation (symbolic matching, combinatorial auctions), task allocation (contract nets)

 

  1.  (March 26): Adjustable autonomy: Decision theoretic approaches, strategies in adjustable autonomy, MDPs [Janusz Marecki will give this lecture]

 

  1. (March 31): Agent modeling I: Symbolic plan recognition, model tracing, prediction

 

  1. (April 2, 7): Agent modeling II: Recursive agent modeling

 

PART III: Adversarial domains, security, safety

 

  1. (April 9, 14): Agent applications to security

·         Praveen Paruchuri, Milind Tambe, Sarit Kraus, Fernando Ordonez Security in multiagent systems by policy randomization International Joint conference on Autonomous agents and Multiagent Systems, 2006

 

  1. (April 16, 21): Project presentation

 

  1. (April 23): Emergence

 

 

  1. (April 28): Review
  2. (April 30): Quiz II

 

 

 Schedule of Assignments and Exams


Dates below are not completely finalized:

·        Homework I handed out: Feb 6, 2008 (10%)

·        Midterm projects due: February 25, 2008 (10%)

·        Quiz I: March 5, 2008  (25%)

·        Panel discussion: March 12, 2008 (5%)

·        Homework II: April 7, 2008 (10%)

·        Project presentation: April 16, April 21 (25%)

·        Final Quiz: April 30, 2008  (15%)