Course detail

Game Theory

FIT-THEAcad. year: 2017/2018

The course deals with Mathematical game theory which is oftenly called the Theory of interactive decision making. The game theory became a popular tool for analysing of intelligent entities in many situations of competition or cooperation. This theory is being commonly applied in area of control, economic models, psychology, sociology, foreign affairs, evolutionary biology and informatics too. By computer science point of view, the game theory is an extension to artificial intelligence with algorithms of decision making, competing and bargaining. This also relates to multi-agent approaches. Games will be treated as models of real or fictitious situations with attributes of intelligence and competition. Students will go through basic terminology of games by the mechanism of their playing (sequential, strategic), by distribution of payoffs in a game (zero/nonzero sum games), by possible cooperation of players (cooperative, non-cooperative) and also by state of information in a game (complete/incomplete information). After the introduction, the games will be extended with possible repetition of moves (repeated games) and its effect to players behavior. In the second part of the semester, we will pay attention to game applications, mechanism design, auctions, social choice, economic and market models and others.

Language of instruction

Czech

Number of ECTS credits

Mode of study

Not applicable.

Guarantor

Ing. Martin Hrubý, Ph.D.

Department

Department of Intelligent Systems (UITS)

Learning outcomes of the course unit

Students will get a wide knowledge of game theory and a plenty of its applications in engineering and social sciences. When passed the course, the students will be able to create a simple model of given game situation and predict its probable future evolution.

In more general level, the study of rational decision making give a certain skills of problem analysis, selecting possible strategies and actions leading to its solving, assigning some utility to the strategies and finally, to accept a best decision in that situation. Mathematical game models also present clearly solutions to many problems in every day life. Moreover, the course introduces and plenty of applications of the computer science to natural and social sciences.

Prerequisites

Students should have a basic knowledge of discrete mathematics, algebra and mathematical analysis, as they are basic tools to describe the studied problems. Basics of artificial intelligence and computer modelling are also required.

Co-requisites

Not applicable.

Planned learning activities and teaching methods

Not applicable.

Assesment methods and criteria linked to learning outcomes

Students have to release a individually made project which will be evaluated at least with one half of points (20 points from 40).

Course curriculum

Syllabus of lectures:

Introduction, history of game theory, motivations to its study, theory of choice, basic terminology, basic classification of games, information in a game.
Two player games with zero-sum payoffs: concept, saddle point, minimax theorem.
Two player games with nonzero-sum payoffs: concept, strategy dominance, Nash equilibrium in pure and mixed strategies, basic algorithms to find the Nash equilibrium.
Mathematical methods in nonzero-sum games: proof of Nashe's lemma of equilibrium existence in games with finite sets of strategies, algorithms to compute the equilibrium, graphical solution to games, linear programming.
Sequential game with perfect/imperfect information: concept, applications, Stackelberg equilibrium, backward induction.
Cooperative games and bargaining: presumptions for possible cooperation, bargaining in nonzero-sum games, Nash bargaining solution.
Repeated games: concept (finite/infinite number of repetitions), solution. Applications of repeated games. Effect of repetitions to players behavior.
Mechanism design: introduction to Mechanism design. Choice under uncertainty.
Social choice, public voting: Arrow's paradox, mechanisms of voting.
Auctions: study of rationality in auctions (mechanism with money). Business applications.
Correlated equilibrium: effect of correlation to rational behavior, definition of correlated equilibrium and its relation to Nash equilibrium. Computing of correlated equilibria, applications.
Evolutionary biology: strategic behavior in population of many entities, evolutionary stable strategy, case studies in the nature.
Applications in economics and engineering: basic solution of oligopoly in analytic and numerical manner, nontrivial case study and its analysis. Application of game theory in computer networks. Applications in psychology, sociology and foreign affairs.

Syllabus - others, projects and individual work of students:

Study - detail reading of given scientific paper and its analysis.
Implementation - implementation of a given algorithm.
Applications - a case-study and its model.

Work placements

Not applicable.

Aims

The THE course is going to give the students certain education in area of rational strategic decision making in conflict situations, to learn them creating models of such situations, to analyze the situations through the models and in some cases to predict future evolution of the modeled systems. The course extends the education of artificial intelligence with strategic decision making. Applications and use will be oriented to the computer science (control, decisions, safety and security, games, networking) and also to social sciences like economics, sociology and political sciences.

Specification of controlled education, way of implementation and compensation for absences

Individual project and final exam. The final exam has two alternativies.
The minimal number of points which can be obtained from the final exam is 20. Otherwise, no points will be assigned to a student.

Recommended optional programme components

Not applicable.

Prerequisites and corequisites

Not applicable.

Basic literature

různí autoři: Classics in Game Theory, edited by Harold W. Kuhn, Princetown University Press, 1997 Cesa-Bianci, N., Lugosi, G.: Prediction, Learning, and Games, Cambridge University Press, 2006 Shubik, M.: Game Theory in the Social Sciences: Concepts and Solutions, MIT Press, 1984 Dresher, M.: The Mathematics of Games of Strategy, Theory and Applications, Dover Publications, 1981 McCarty, N., Mierowitz, N.: Political Game Theory: An Introduction, Cambridge University Press, 2007 různí autoři: Algorithmic Game Theory, edited by Noam Nisan, Cambridge University Press, 2006 Osbourne, M.J., Rubinstein, A.: A Course in Game Theory, MIT Press, 1994 Fudenberg, D., Tirole, J.: Game Theory, MIT Press, 1991 Dorfman, R., Samuelson, P.A., Solow, R. M.: Linear Programming and Economic Analysis, Dover Publications, 1986 Schelling, T. S. : The Strategy of Conflict, Harvard Press, 1980 Dugatkin, L., Reeve, H.: Game Theory and Animal Behavior, Oxford University Press, 1988 Morrow, J.: Game Theory for Political Scientists, Princeton University Press, 1994 Kreps, D.: Game Theory and Economic Modelling, Oxford University Press, 1990 von Neumann, J., Morgenstern, O.: Theory of Games and Economic Behavior, Princeton University Press, 1944 Mailath, G., Samuelson, L.: Repeated Games and Reputations, Oxford University Press, 2006 Krishna, V.: Auction Theory, Elsevier, 2002 Gintis, H.: Game Theory Evolving, Princeton University Press, 2000 Miller, J.: Game Theory at Work, McGraw-Hill, 2003 Straffin, P.D.: Game Theory and Strategy, The Mathematical Association of America, 2003 Rasmunsen, E.: Games and Information, Blackwell Publishing, 2007

VUT

Faculties

University Institutes

Parts

Game Theory