SYSTEMS THEORY  (2008)

A very short introduction to foundations of multi-agent systems:
Game  theory, Epistemic Logic, Process theory, and Agent-based programing


Department of Mathematical Sciences

Rasool Ramezanian




Outline
Introduction

GAME THEORY
  1. Lecture G1
  2. Lecture G2
  3. Lecture G3
  4. Lecture G4
  5. Lecture G5

EPISTEMIC LOGIC
  1. Lecture E1
  2. Lecture E2
  3. Lecture E3
  4. Lecture E4
  5. Lecture E5

PROCESS THEORY
  1. Lecture P1
  2. Lecture P2
  3. Lecture P3
  4. Lecture P4

 AGENT-BASED PROGRAMING
  1. L1: BDI
  2. L2: 3APL

I think any science needs a calculus to make it sound and exact. A calculus for a science consists of a language for specifying, and a set of rules for verifying the problems of that science. I believe that a calculus for Theoretical Social Science (economics, political science, law, social psychology, and ...) is Formal Logic (game theory, epistemic logic, deontic logic, and etc).

    Assume a flock of birds immigrating from the Caspian see to the Persian golf. They immigrate in a very neat shape. It is possible to assume that there is a leader in front, and all other birds notice their direction of flying looking the leader; if the leader swerves to left (right), all the birds swerve to that direction as well, and it is why the immigration looks very neat. Another possible assumption is that 
1. each bird produces a smell, 2. the smell produced by the birds outspread in the air, and 3. each bird tries to move to a position where it can detect a specific amount of smell. 
    In this way, the neatness is not a consequence of an ultimate bird, it is a consequence of all birds. Therefore, we may consider two different kinds of controlling factor for an orderly society (of human-beings, animals, robots and etc): 1. ultimate controlling and 2. multi-agent controlling.

    In a multi-agent system, each agent is a rational individual with specific beliefs and obligations, who communicates with other agents in parallel. Also as time passes, agents may develop (evolution), interact and cooperate with each other.
    The rationality of agents is discussed in game theory, where agents aim to maximize their utilities. The beliefs and obligations of agents are specified through epistemic logic, and we consider the concurrency and parallel communication through process theory.


















توجه: پروژه ها بطور دقيق توسط  استاد درس مورد بازبيني قرار نگرفته اند خطاهاي احتمالي ممكن است



 


Course Grade: Your course grade will be determined by a combination of the following factors: 1. Students talk (20%), 2. Students tasks (20%), and 3. final exam (60%).  Students tasks will typically be handed in the lecture notes, and should be back in a week. Late tasks will not be accepted unless a  permission is obtained in advance.