CMSC 477/677Agent Architectures andMulti-Agent Systems

UMBCProf. Marie desJardins

Spring 2005

Course information

Prof desJardins ITE 337, x53967, mariedj@cs.umbc.edu

Class mailing list agents-class@listproc.umbc.edu To subscribe, send email to

listproc@listproc.umbc.edu with the line: subscribe agents-class Your Name

Today’s overview

Class structure and policies What’s an agent? Agent exercise Next class

Class structure: Syllabus

Course page: http://www.cs.umbc.edu/courses/graduate/677/spring05/http://www.cs.umbc.edu/courses/undergraduate/477/spring05/

Class structure: 477 vs. 677

Slightly different weights for assignments Two problem sets for graduate students Agent architectures project: Graduate students must

do a more in-depth analysis, relating their findings to the research literature

MAS project: Graduate students must include an experimental research component, and submit a research design

In general, graduate students are expected to show greater depth in their analysis and synthesis of ideas

Class structure: Participation

This is a discussion class Reading must be done in advance Participation counts—a lot

40/35% of grade is related to class participation Class discussion (30/25%)

Do you attend class? Are you prepared? Have you done the reading? Have you

thought about the discussion questions? Do you contribute to the discussion with insightful questions

and comments? Paper summaries (5%) Discussion leaders (5%)

Class structure: Agent architecture project Agent architecture project: 20/15% of grade

Download one of the architectures we learn about Apply the architecture to a domain of your choice

Deadlines: Proposal due Feb. 17 (5% of project grade) Report due Mar. 17 (70% of grade) Demonstration week of Mar. 14 (25% of grade)

Class structure: MAS paper/presentation MAS paper/presentation: 25% of grade

Students will select a topic to study in greater depth, write a paper, and give a presentation on that topic. 477: can focus primarily on one or two recent research papers 677: can focus on one or two main papers, but should also include a

bibliography of 5-10 (more is OK) papers on the topic, and a significant discussion/analysis of the work in that area.

Proposal and bibliography due Apr. 12 (10% of project grade) Draft report due May 5 (5%) Presentation on May 3, 5, 10, 12(?), or 19 (20%)

additional days if needed: May 13 and/or May 6 Final report due May 19 (65%)

Paper review (of another student’s paper): 5% of grade

MAS competition

Multi-agent game (trading agents?) project: 10% of grade In-class competition – probably May 12 Short report describing design and performance of

agent

Policies

Grading and academic honesty Plagiarism, citations

Original passage: I pledge allegiance to the flag of the United States

of America, and to the republic for which it stands, one nation, indivisible, with liberty and justice for all.

Unacceptable summary: I promise loyalty to the United States flag, and to

the country for which it stands, one nation, with freedom and fairness for all.

Plagiarism exercise

Plagiarism exercise II

Original passage: I pledge allegiance to the flag of the United States

of America, and to the republic for which it stands, one nation, indivisible, with liberty and justice for all.

Acceptable summary: I promise to be loyal to the United States flag and

to the USA itself: One united country that provides basic rights such as liberty and justice to all citizens.

What’s an agent?

Weiss, p. 29 [after Wooldridge and Jennings]: “An agent is a computer system that is situated in some

environment, and that is capable of autonomous action in this environment in order to meet its design objectives.”

Russell and Norvig, p. 7: “An agent is just something that perceives and acts.”

Rosenschein and Zlotkin, p. 4: “The more complex the considerations that [a] machine

takes into account, the more justified we are in considering our computer an ‘agent,’ who acts as our surrogate in an automated encounter.”

What’s an agent? II

Ferber, p. 9: “An agent is a physical or virtual entity

a) Which is capable of acting in an environment,b) Which can communicate directly with other agents,c) Which is driven by a set of tendencies…,d) Which possesses resources of its own,e) Which is capable of perceiving its environment…,f) Which has only a partial representation of this environment…,g) Which possesses skills and can offer services,h) Which may be able to reproduce itself,i) Whose behavior tends towards satisfying its objectives, taking

account of the resources and skills available to it and depending on its perception, its representations and the communications it receives.”

OK, so what’s an environment? Isn’t any system that has inputs and outputs

situated in an environment of sorts?

What’s autonomy, anyway?

Jennings and Wooldridge, p. 4: “[In contrast with objects, we] think of agents as

encapsulating behavior, in addition to state. An object does not encapsulate behavior: it has no control over the execution of methods – if an object x invokes a method m on an object y, then y has no control over whether m is executed or not – it just is. In this sense, object y is not autonomous, as it has no control over its own actions…. Because of this distinction, we do not think of agents as invoking methods (actions) on agents – rather, we tend to think of them requesting actions to be performed. The decision about whether to act upon the request lies with the recipient.”

Is an if-then-else statement sufficient to create autonomy?

So now what?

If those definitions aren’t useful, is there a useful definition? Should we bother trying to create “agents” at all?

Next class

Reading: Wooldridge Chapter 1; Wooldridge & Jennings 1995

Overview by Dr. dJ

Tuesday reading: Wooldridge Chapter 3; Levesque et al. 1997

Discussion leaders!

Multi-agent exercise

Getting to know you... getting to know all about you... (or at least your capabilities...)

After-action reviewor post-mortem, as the case may be…

What was the task completion rate? How many agents participated in successful

teams? Who was more successful – agents who led

teams, or agents who participated on teams? Any particularly successful (or unsuccessful)

strategies for forming teams? What’s hard about this problem?