The Wisdom of Networked and Evolving Agents
|
|
- Peter Bell
- 5 years ago
- Views:
Transcription
1 The Wisdom of Networked and Evolving Agents How mutual beneficial rules can be built and sustained without knowing the motives of others, and without sharing goals. Akira Namatame National Defense Academy,Japan Artificial Economics 06 (Aalborg)
2 Social Intelligence <Intelligence is valuable in social contexts>. Economic theory presupposes rationality that rely on the existence of intelligence, but why would such intelligence has evolved? Two hypotheses have been considered as explanations for this question. (1) The ecological hypothesis: The demands of the non-human environment were a key impetus to higher intelligence. (2) The social intelligence hypothesis: A key factor is interactions with other human beings. :The Red Queen effect (co-evolution) in biology was introduced to explain why species have been evolved to higher intelligence. 1 - Artificial Economics 06 (Aalborg)
3 Basic Issues [1] (1) Why do we need to focus on interactions? Global stability despite many local interactions It is remarkable that in almost all cases interactions play a vital role, yet many aggregate outcomes are not changed too seriously by them. (Philip W. Anderson: in his Nobel Prize lecture, 1977) (2) Endless and meaningless co-evolution? The Red Queen principle (arms race) The more things change, the more they stay the same <van Valen (evolutionary biologists), 1973> 2 - Artificial Economics 06 (Aalborg)
4 Flocking Behaviors What is the mechanism behind flocking behavior? 3 - Artificial Economics 06 (Aalborg)
5 Relationship Rather Than Rationality BOID Model: Craig W. Reynolds (1994) <Frocking Behavior consists of three relational rules> Cohesion: Head for the perceived center of mass of the neighbors. Separation: Don't get too close to any object. Alignment: Try to match the speed and direction of nearby neighbors. Cohesion Separation Alignment 4 - Artificial Economics 06 (Aalborg)
6 Basic Issues [2] Self-interest vs. Mutual Benefit Self-interested agents are often faced with the dilemma of acting in their own interest or pursuing a more cooperative action in social contexts. The prevailing orthodoxy in the evolutionary theory is that the better individual (or gene) is the unit of selection. :Group solidarity would work for a desirable outcome for mutual benefit, but evolutionary dynamics based on natural selection can work against it. A challenging task is to identify conditions under which agents are more coordinated than they are acting in their own interest. 5 - Artificial Economics 06 (Aalborg)
7 Madness of Crowds vs. Wisdom of Crowds (1) Researches on collectives of agents Social sciences: <Crowd psychology> herding, cascade, group think, Game theory: <the problem of a coordination failure> The existence of externalities lead to coordination failure and results in Pareto-inferior outcomes. Computer science: <the price of anarchy> Selfish behavior may not achieve full efficiency at the collective level. 6 - Artificial Economics 06 (Aalborg)
8 The Price of Anarchy (1) Coordination versus competition Equilibria of selfish agents are not efficient Pareto optimal is usually not Nash equilibrium Price of anarchy: Quantify inefficiency in terms of a global objective Price of Anarchy = objective function value at NEP optimal objective function value 7 - Artificial Economics 06 (Aalborg)
9 The Price of Anarchy (2) Agent A Agent B Pure strategy Nash equilibria include (H,H) and (M, M), but not (L,L) The efficient outcome is (L, L) H M L A modified coordination game H 5,5 M 0,0 L 12,0 0,0 2,2 0,0 0,12 0,0 10,10 Price of Anarchy=5(or 2)/10=1/2 (or1/5) -How can we improve the price of anarchy without the central authority? -What mechanisms one can use to improve the overall system performance? 8 - Artificial Economics 06 (Aalborg)
10 Madness of Crowds vs. Wisdom of Crowds (2) The Wisdom Of Crowds, Surowiecki, Random House (2004) A large collection of people are smarter than an elite few. (Example: Wikipedia) Surowiecki suggests new insights regarding how our social and economic activities should be organized. The wisdom of crowds emerges only under the right conditions, which are : diversity, : independence, : decentralization, and : aggregation. Network topology of interactions also play a vital role. 9 - Artificial Economics 06 (Aalborg)
11 Concept of Social Games Networked agents repeated play pair-wise games: Coordination game Nash demand game Rock-scissors-paper game Agents evolve their strategies Best-response strategy learning has been the focus of much research. However, this approach can yield undesirable outcomes. We seek a proper learning model leading agents to desirable outcomes Artificial Economics 06 (Aalborg)
12 Effects of Network Topology It is important to interact with the right peoples. It is not what you know, it s who you know that account. Payoff: Internal enforcement Lattice network AGENT Network topology: External enforcement Small-world network: Rewiring with probability p=0.5 Random network: Rewiring with probability p= Artificial Economics 06 (Aalborg)
13 Fixed Rule vs. Learnable Rule Learning: Memory-based strategy choice : Memory A t (strategy at time t) Is a short memory enough or a longer memory may be necessary in more complex games? : Reactionary strategy with history one: ( h t-1 ) A t : Finite history of k: { ( h t-k,, h t-2, h t-1 ) } A t Agents play the game repeatedly by evolving a rule rather than by a fixed rule such as TFT. bit past strategy Own Opp strategy at t # # # # Crossover 12 - Artificial Economics 06 (Aalborg)
14 A Coordinated Attack Problem Agent B S Agent A 1 (attack) S 2 (retreat) S 1 (attack) S 2 (retreat) Coordination Game The inferior strategy S 2 is likely to be selected. The basin of attraction of the retreat is 9 times as large as that of the attack. θ= Agents need to decide to attack or retreat. If they both attack, they win. However, one sided attack would result in losing the battle. (S 1, S 1 ): Pareto-dominates (S 2, S 2 ), (S 2, S 2 ): risk-dominates (S 1, S 1 ) S1 θ=2/3 S2 Pareto-dominance θ=0.1 risk-dominance From the point of view of social welfare this is a shame because agents are better off at the joint attack equilibrium Artificial Economics 06 (Aalborg)
15 Simulation Results: Coordination Game generation 200 generation 300 generation 2000 Lattice network average 0.99 average 1.0 average 0.98 Small-world network average 1.0 average 1.0 average 1.0 Random network average 0.0 average Artificial Economics 06 average (Aalborg) 0.0
16 What did Agents Learn through Co-evolution? bit past strategy Own Opp strategy at t # # # # The rules of 2,500 agents are aggregated into a few rules. S 1 : 0 S 2 : 1 Array location 1: Initial strategy 2: Choice when outcome is (0, 0) 3: Choice when outcome is (0, 1) 4: Choice when outcome is (1, 0) 5: Choice when outcome is (1, 1) random network lattice network small-world network 15 - Artificial Economics 06 (Aalborg)
17 The Phase Diagram of the Plays of Two Agents lattice network small-world network converges to 00 : Pareto-optimal outcome Initial strategy random network converges to 11 :risk-dominat outcome 16 - Artificial Economics 06 (Aalborg)
18 Evolutionary Bargaining: Sharing Game Agent A Agent B H = 7 M = 5 L = 3 The payoff matrix of a sharing game H = 7 0,0 M = 5 0,0 L = 3 7,3 0,0 5,5 5,3 3,7 3,5 3,3 There are three Nash equilibriums in pure strategy with one equitable (M, M), and two inequitable ones (H,L), (L,H) H Best reply: L Belief of agent «i» (Phan, 2005) M Best reply: M Best reply: H L 17 - Artificial Economics 06 (Aalborg)
19 Emergence of Self-reinforcing Classes (Axtell, 2001) Selective interaction: Agents can distinguish each other using tags. Classes formation: Agents behave fair (M) to agents within the same group but behave unfair between groups <Left-hand side> Both black and grey agents are fair within their group (Play M) <Right-hand side>black agents have beliefs in behavior of grey agents, and their best response is to play L (70%). Grey agents have beliefs in behavior of blacks. agents, and their best response is to play H, and therefore unfair outcome (H, L) is realized. H H M Within-group L M (Phan,2005) Between-group L 18 - Artificial Economics 06 (Aalborg)
20 Simulation Results: Sharing Game (i) The average payoff per agent : Highest payoff Average Payoff (ii) the ratio of each strategy S ( M 2 ) S 1( Rock ) H=7 M=5 L= 3 H = 7 0,0 M = 5 0,0 L = 3 7,3 0,0 5,5 5,3 3,7 3,5 3,3 All agents become to choose M and realize both efficient and fair outcome 19 - Artificial Economics 06 (Aalborg)
21 Repeated Games and the General Theory of Efficiency Economics recognizes three general ways to achieve efficiency: Competition: efficient resources allocation via markets Contracts/mechanism design: as a means to enforce efficient outcomes Repeated games: provide a formal framework to examine why selfinterested agents manage to coordinate in a long-term relationship 20 - Artificial Economics 06 (Aalborg)
22 <The Folk Theorem> The Folk Theorem in Repeated Games In a repeated interaction, any mutually beneficial outcome can be sustained in an equilibrium. <Criticism of The Folk Theorem> The theory of repeated games does not provide a criterion for equilibrium selection The majority of works on repeated games are symmetric games. The theory of repeated games is rather silent for asymmetric situations Artificial Economics 06 (Aalborg)
23 Generalized Rock-Scissors-Paper Games λ=2: Conventional R-S-P game λ>2: Dispersion games A Nash equilibrium strategy: (S 1 : 1/3, S 2 : 1/3, S 3 : 1/3) The expected payoff at Nash equilibrium : (λ+1)/3 Pareto-efficiency (average payoff): λ/2 Inequity at Pareto-efficiency : A lucky agent gets λ, and an unlucky agent gets nothing Artificial Economics 06 (Aalborg)
24 Characteristics of RSP Game Group solidarity would work for a Pareto-efficient outcome for mutual benefit, but evolutionary dynamics, which is based on natural selection, can work against it. The RSP game is not stable in the evolutionary dynamics. :However, with slight modification of the payoff of the original RSP game, the Nash equilibrium becomes a stable attractor Artificial Economics 06 (Aalborg)
25 Coupling Rules with Three Strategies Previous Strategy Next Strategy Own Opponent 0 0 # 0 1 # 0 2 # 1 0 # 1 1 # 1 2 # 2 0 # 2 1 # 2 2 # Coupled rule Strategy choice is driven by the outcomes of the joint actions Learnable coupled rule Agents update the coupling rule by changing value of #. 0:Rock, 1: Scissor, 2: Paper #: 0, 1 or Artificial Economics 06 (Aalborg)
26 Simulation Results (1): λ=2 max Error rate: 0% S 2 ( Scissor ) max min average (i) The average payoff per agent Error rate: 10% (ii) the ratio of each strategy max S 2 ( Scissor ) min S 1 ( Rock 25 - Artificial Economics 06 (Aalborg) )
27 What Did Agent Learn through Co-evolution? λ=2 (Error rate: 0%) The rules of 400 agents are aggregated into 67 types Learned coupling rules # # # # 0 # Nash equilibrium: 1 Pareto-efficiency: 1 The phase diagram of the plays between two agents Initial strategy 26 - Artificial Economics 06 (Aalborg)
28 Simulation Results(2): λ=10 : Highest payoff Nash equilibrium: 3.3 Pareto-efficiency: (i) The average payoff per agent Error rate: 0% Average Payoff 2.8 : Lowest Payoff (ii) the ratio of each strategy S 2( Scissor ) S 1( Rock S 3( Paper The strategy population is close to Nash equilibrium ) ) Error rate: 10% 4.0 S 1( Rock ) S 2( Scissor ) S 3( Paper ) 27 - Artificial Economics 06 (Aalborg)
29 What Did Agents Learn through Co-evolution? λ=10(error rate: 10%) The rules of 400 agents are aggregated into 8 rules with some common values Rule Type Number of Agents with the Same Rule Rule Type Rule Type Rule Type Rule Type 5 Rule Type 6 Rule Type 7 Rule Type Artificial Economics 06 (Aalborg)
30 Phase Diagram: Two Agents with the Same Rule If the two agents of having the same rule play, they absorb in the drawing cycle. draw win lose Initial State 11 draw win lose draw win lose 29 - Artificial Economics 06 (Aalborg)
31 Phase Diagram: Agents with the Different Rules If the two agents of having the different rules play, they absorb in the limiting cycle. draw Winning phase win Losing phase lose Initial State 11 draw win Limiting cycle lose draw win lose 30 - Artificial Economics 06 (Aalborg)
32 Emergence of Turn-taking Behavior By visiting all efficient outcomes alternatively under turn-taking behavior, the agents could achieve desirable outcomes in efficiency and fairness Artificial Economics 06 (Aalborg)
33 Concluding Remarks (1): The inverse problem of networked agents <The forward problem> Desired collectives : Analysis,Explanation, Understanding <The inverse problem> :Infer micro behavioral rules of agents : Design micro-rules through co-evolutionary learning Emergent Inverse problem behavior Forward problem Interacting agents with micro-motives 32 - Artificial Economics 06 (Aalborg)
34 Concluding Remarks (2): ABM vs. DE: What are the differences? Agent Based models (ABM) are widespread and many exciting applications, but lots of hype, not enough understanding of when ABM adds value. Replicator dynamics (DE) is the most thoroughly studied dynamic model for evolution with the more successful strategies being imitated more often. What are the differences between ABM and Differential Equations (DE) methods? 33 - Artificial Economics 06 (Aalborg)
35 Rule Rewards Learning Algorithm Meta-strategy Concluding Remarks (2): ABM vs. DE: What are the differences?(cont.) Best-response strategy learning has been the focus of much research. However, this approach can yield undesirable outcomes. Replicator dynamics converge to Nash equilibria which are not always desirable outcomes. Observations, Sensations World, State strategy <Learning relationship> -Agents should learn a better way to be coupled with others -ABM is the only way for the study of relationship 34 - Artificial Economics 06 (Aalborg)
36 Concluding Remarks [3]: Collective Norm Development Social norms are rules that are understood by members of a society, and that guide and constrain their behavior without the force of laws. Emergence and sustainability of social norms as core issues for social sciences. Social norms emerge out of social interaction and they are internalized in networked agents. Social norm Agent Internalization of social norm shared rule Agent preference 35 - Artificial Economics 06 (Aalborg)
37 Open Problems: Smart Agents Using ICT (ICT:Information/Communication Technology) Congestion control receive much attention What wisdom for avoiding congestion will be emerged in networked smart agents equipped with ICT? 36 - Artificial Economics 06 (Aalborg)
38 Open Problems(cont.): Smart Networks Design by Agents Smart Business Networks,Ubiquitous Web of Services, How mutual trust can be built and sustained without knowing the motives of others, and without sharing goals. Smart Networks Design with Genetics Approach 37 - Artificial Economics 06 (Aalborg)
39 New book: World Scientific, Artificial Economics 06 (Aalborg)
40 Thank you!! and Question time 39 - Artificial Economics 06 (Aalborg)
UNIK Multiagent systems Lecture 5. Non-cooperative game theory. Jonas Moen
UNIK4950 - Multiagent systems Lecture 5 Non-cooperative game theory Jonas Moen Highlights lecture 5 Non-cooperative game theory* Classification of game theory Utility of self-interested agents Strategic
More informationEconomics II - October 27, 2009 Based on H.R.Varian - Intermediate Microeconomics. A Modern Approach
Economics II - October 7, 009 Based on H.R.Varian - Intermediate Microeconomics. A Modern Approach GAME THEORY Economic agents can interact strategically in a variety of ways, and many of these have been
More informationTopics in ALGORITHMIC GAME THEORY *
Topics in ALGORITHMIC GAME THEORY * Spring 2012 Prof: Evdokia Nikolova * Based on slides by Prof. Costis Daskalakis Let s play: game theory society sign Let s play: Battle of the Sexes Theater Football
More informationGame theory (Sections )
Game theory (Sections 17.5-17.6) Game theory Game theory deals with systems of interacting agents where the outcome for an agent depends on the actions of all the other agents Applied in sociology, politics,
More information"Solutions" to Non-constant Sum Games
Unit 3 GAME THEORY Lesson 30 Learning Objective: Analyze two person non-zero sum games. Analyze games involving cooperation. Hello students, The well-defined rational policy in neoclassical economics --
More informationLecture 21: Strategic Interaction and Game Theory
Lecture 2: Strategic Interaction and Game Theory EC DD & EE / Manove Strategic Interaction p What types of firms are most likely to engage in costly rent-seeking? EC DD & EE / Manove Monopoly>Rent Seeking
More informationLecture 1 - Introduction
Lecture 1 - Introduction 14.03 Spring 2003 1 Introduction What is economics? A social science of human decision-making, like psychology, sociology, anthropology. What are the objectives of economics? 1.
More informationAlgoritmi Distribuiti e Reti Complesse (modulo II) Luciano Gualà
Algoritmi Distribuiti e Reti Complesse (modulo II) Luciano Gualà guala@mat.uniroma2.it www.mat.uniroma2.it/~guala Algorithmic Game Theory Algorithmic Issues in Non-cooperative (i.e., strategic) Distributed
More informationCS 486/686 Lecture 17 2-Player Normal-Form Games 1. Determine dominant-strategy equilibria of a 2-player normal form game.
CS 486/686 Lecture 17 2-Player Normal-Form Games 1 Learning goals: By the end of the lecture, you should be able to Determine dominant-strategy equilibria of a 2-player normal form game. Determine pure-strategy
More informationAutonomous Agents and Multi-Agent Systems* 2015/2016. Lecture Reaching Agreements
Autonomous Agents and Multi-Agent Systems* 2015/2016 Lecture Reaching Agreements Manuel LOPES * These slides are based on the book by Prof. M. Wooldridge An Introduction to Multiagent Systems and the online
More informationSupplimentary material for Research at the Auction Block: Problems for the Fair Benefits Approach to International Research
Supplimentary material for Research at the Auction Block: Problems for the Fair Benefits Approach to International Research Alex John London Carnegie Mellon University Kevin J.S. Zollman Carnegie Mellon
More informationPRICE OF ANARCHY: QUANTIFYING THE INEFFICIENCY OF EQUILIBRIA. Zongxu Mu
PRICE OF ANARCHY: QUANTIFYING THE INEFFICIENCY OF EQUILIBRIA Zongxu Mu The Invisible Hand Equilibria and Efficiency Central to free market economics The Wealth of Nations (Smith, 1776) led by an invisible
More information6.896: Topics in Algorithmic Game Theory
6.896: Topics in Algorithmic Game Theory vol. 1: Spring 2010 Constantinos Daskalakis game theory what society we won t sign study in this class I only mean this as a metaphor of what we usually study in
More informationGames, Auctions, Learning, and the Price of Anarchy. Éva Tardos Cornell University
Games, Auctions, Learning, and the Price of Anarchy Éva Tardos Cornell University Games and Quality of Solutions Rational selfish action can lead to outcome bad for everyone Tragedy of the Commons Model:
More informationPrice of anarchy in auctions & the smoothness framework. Faidra Monachou Algorithmic Game Theory 2016 CoReLab, NTUA
Price of anarchy in auctions & the smoothness framework Faidra Monachou Algorithmic Game Theory 2016 CoReLab, NTUA Introduction: The price of anarchy in auctions COMPLETE INFORMATION GAMES Example: Chicken
More informationECS 253 / MAE 253, Lecture 13 May 10, I. Games on networks II. Diffusion, Cascades and Influence
ECS 253 / MAE 253, Lecture 13 May 10, 2016 I. Games on networks II. Diffusion, Cascades and Influence Summary of spatial flows and games Optimal location of facilities to maximize access for all. Designing
More information5/2/2016. Intermediate Microeconomics W3211. Lecture 25: Recap 2. The Story So Far. Organization for the Week. Introduction
1 Intermediate Microeconomics W3211 Lecture 25: Recap 2 Introduction Columbia University, Spring 2016 Mark Dean: mark.dean@columbia.edu 2 The Story So Far. 3 The Story So Far. 4 Topic Topic 1 The Consumer
More informationIntro to Algorithmic Economics, Fall 2013 Lecture 1
Intro to Algorithmic Economics, Fall 2013 Lecture 1 Katrina Ligett Caltech September 30 How should we sell my old cell phone? What goals might we have? Katrina Ligett, Caltech Lecture 1 2 How should we
More informationChapter 9: Static Games and Cournot Competition
Chapter 9: Static Games and Cournot Competition Learning Objectives: Students should learn to:. The student will understand the ideas of strategic interdependence and reasoning strategically and be able
More informationOligopoly: How do firms behave when there are only a few competitors? These firms produce all or most of their industry s output.
Topic 8 Chapter 13 Oligopoly and Monopolistic Competition Econ 203 Topic 8 page 1 Oligopoly: How do firms behave when there are only a few competitors? These firms produce all or most of their industry
More informationAlgoritmi Distribuiti e Reti Complesse (modulo II) Luciano Gualà
Algoritmi Distribuiti e Reti Complesse (modulo II) Luciano Gualà guala@mat.uniroma2.it www.mat.uniroma2.it/~guala Algorithmic Game Theory Algorithmic Issues in Non-cooperative (i.e., strategic) Distributed
More informationHarvard University Department of Economics
Harvard University Department of Economics General Examination in Microeconomic Theory Spring 00. You have FOUR hours. Part A: 55 minutes Part B: 55 minutes Part C: 60 minutes Part D: 70 minutes. Answer
More informationCo-evolution of strategies and update rules in the prisoner s dilemma game on complex networks
Co-evolution of strategies and update rules in the prisoner s dilemma game on complex networks Alessio Cardillo Department of Physics of Condensed Matter University of Zaragoza and Institute for Biocomputation
More informationLecture 21: Strategic Interaction and Game Theory
Lecture : Strategic Interaction and Game Theory EC DD & EE / Manove Strategic Interaction p EC DD & EE / Manove Clicker Question p Strategic Interaction In perfectly competitive markets (like the market
More informationSI Game Theory, Fall 2008
University of Michigan Deep Blue deepblue.lib.umich.edu 2008-09 SI 563 - Game Theory, Fall 2008 Chen, Yan Chen, Y. (2008, November 12). Game Theory. Retrieved from Open.Michigan - Educational Resources
More informationAn Introduction to Game Theory
OXFORD PROSPECTS 3 st July and 6 th August 8 Slides are available at: users. ox. ac. uk/ ~sedm375/ gametheory. pdf richard.povey@hertford.ox.ac.uk, richard.povey@st-hildas.ox.ac.uk Overview - Purpose of
More informationCMSC 474, Introduction to Game Theory Analyzing Normal-Form Games
CMSC 474, Introduction to Game Theory Analyzing Normal-Form Games Mohammad T. Hajiaghayi University of Maryland Some Comments about Normal-Form Games Only two kinds of strategies in the normal-form game
More informationMiscomputing Ratio: The Social Cost of Selfish Computing
Miscomputing Ratio: The Social Cost of Selfish Computing Kate Larson and Tuomas Sandholm Computer Science Department Carnegie Mellon University 5000 Forbes Ave Pittsburgh, PA 15213 {klarson,sandholm}@cs.cmu.edu
More informationTutorial: Introduction to Game Theory
July, 2013 Tutorial: Introduction to Game Theory Jesus Rios IBM T.J. Watson Research Center, USA jriosal@us.ibm.com Approaches to decision analysis Descriptive Understanding of how decisions are made Normative
More informationGatti: A gentle introduction to game theory and crowding games. Outline. Outline. Nicola Gatti
gentle introduction to game theory and crowding games Nicola Gatti Dipartimento di Elettronica e Informazione Politecnico di Milano Italy Outline Examples Game theory groundings Searching for a Nash equilibrium
More informationModels of Language Evolution
Models of Language Evolution Session 4: 2014/11/19 Organizational Matters 22.10 Language Evolution - Overview 29.10 Language Evolution - Protolanguage 12.11 Models of Language Evolution 19.11 26.11 Evolutionary
More informationECO 5341 Strategic Behavior Lecture Notes 1
ECO 5341 Strategic Behavior Lecture Notes 1 Saltuk Ozerturk Southern Methodist University Spring 2016 (Southern Methodist University) Introduction to Game Theory Spring 2016 1 / 14 What is game theory?
More informationWhere are we? Knowledge Engineering Semester 2, Basic Considerations. Decision Theory
H T O F E E U D N I I N V E B R U S R I H G Knowledge Engineering Semester 2, 2004-05 Michael Rovatsos mrovatso@inf.ed.ac.uk Lecture 13 Distributed Rational Decision-Making 25th February 2005 T Y Where
More informationEcon 3542: Experimental and Behavioral Economics Exam #1 Review Questions
Econ 3542: Experimental and Behavioral Economics Exam #1 Review Questions Chapter 1 (Intro) o Briefly describe the early history of market experiments. Chamberlin s original pit market experiments (1948)
More informationAdaptive Properties and Memory of a System of Interactive Agents: A Game Theoretic Approach
Adaptive Properties and Memory of a System of Interactive Agents: A Game Theoretic Approach Roman Gorbunov, Emilia Barakova, Rene Ahn, Matthias Rauterberg Designed Intelligence Group, Department of Industrial
More informationDo not open this exam until told to do so. Solution
Do not open this exam until told to do so. Department of Economics College of Social and Applied Human Sciences K. Annen, Fall 003 Final (Version): Intermediate Microeconomics (ECON30) Solution Final (Version
More informationObjective. Sessions on Economics. Types of Economic Analysis. Session 2
Objective Sessions on Economics Pharm 532 Lou Garrison, Ph.D. April 9-23, 2007 Understand the basic principles of economics as applied to health care and integrate there principles into policy analysis.
More informationChapter Fourteen. Topics. Game Theory. An Overview of Game Theory. Static Games. Dynamic Games. Auctions.
Chapter Fourteen Game Theory Topics An Overview of Game Theory. Static Games. Dynamic Games. Auctions. 2009 Pearson Addison-Wesley. All rights reserved. 14-2 Game Theory Game theory - a set of tools that
More informationCitation for published version (APA): Kopányi, D. (2015). Bounded rationality and learning in market competition Amsterdam: Tinbergen Institute
UvA-DARE (Digital Academic Repository) Bounded rationality and learning in market competition Kopányi, D. Link to publication Citation for published version (APA): Kopányi, D. (2015). Bounded rationality
More informationI. Introduction to Public Goods and Public Goods Problems
I. Introduction to Public Goods and Public Goods Problems A. One of the many postwar innovations in economic theory was the idea of a pure public good. This idea was first clearly stated by Paul Samuelson
More informationTwo articles: Fairness and Retaliation: The Economics of Reciprocity Explaining Bargaining Impasse: The Role of Self-Serving Biases
Two articles: Fairness and Retaliation: The Economics of Reciprocity Explaining Bargaining Impasse: The Role of Self-Serving Biases Behavioural/experimental economics - results mainly based on economic
More informationMERITOCRACY AS A MECHANISM IN THE CONTEXT OF VOLUNTARY CONTRIBUTION GAMES
MERITOCRACY AS A MECHANISM IN THE CONTEXT OF VOLUNTARY CONTRIBUTION GAMES HEINRICH H. NAX (HNAX@ETHZ.CH) COSS, ETH ZURICH MAY 26, 2015 BUT BEFORE WE BEGIN Let us clarify some basic ingredients of the course:
More informationEcon8500_Game_Theory. Name: Class: Date: Multiple Choice Identify the choice that best completes the statement or answers the question.
Name: Class: Date: ID: A Econ8500_Game_Theory Multiple Choice Identify the choice that best completes the statement or answers the question. 1. A common assumption about the player in a game is that a.
More informationFOUR MARKET MODELS. Characteristics of Oligopolies:
FOUR MARKET MODELS Perfect Competition Monopolistic Competition Oligopoly Pure Monopoly Characteristics of Oligopolies: A Few Large Producers (Less than 10) Identical or Differentiated Products High Barriers
More informationGame Theory and Economics
and Economics Game theory is the study of how people behave in strategic situations. Strategic decisions are those in which each person, in deciding what actions to take, must consider how others might
More information1 Competitive Equilibrium
1 Competitive Equilibrium Each household and each firm in the economy act independently from each other, seeking their own interest, and taking as given the fact that other agents will also seek their
More informationAlgorithmic Collusion For IO Reading Group
Algorithmic Collusion For IO Reading Group Chris Doyle, Department of Economics, Warwick University, 14 March 2018 Papers Review paper: Algorithmic Collusion: Problems and Counter-Measures by Ezrachi and
More informationECN 3103 INDUSTRIAL ORGANISATION
ECN 3103 INDUSTRIAL ORGANISATION 5. Game Theory Mr. Sydney Armstrong Lecturer 1 The University of Guyana 1 Semester 1, 2016 OUR PLAN Analyze Strategic price and Quantity Competition (Noncooperative Oligopolies)
More informationFIRST FUNDAMENTAL THEOREM OF WELFARE ECONOMICS
FIRST FUNDAMENTAL THEOREM OF WELFARE ECONOMICS SICONG SHEN Abstract. Markets are a basic tool for the allocation of goods in a society. In many societies, markets are the dominant mode of economic exchange.
More informationWelfare Economics. The Edgeworth Box. The Basic Theorem. Some Basic Assumptions
Welfare Economics The Edgeworth Box The Basic Theorem The basic theorem in welfare economics: A market, exchange, economy will achieve efficient resource allocation. We intend to show the basics of that
More informationIntroduction to Genetic Algorithm (GA) Presented By: Rabiya Khalid Department of Computer Science
Introduction to Genetic Algorithm (GA) Presented By: Rabiya Khalid Department of Computer Science 1 GA (1/31) Introduction Based on Darwin s theory of evolution Rapidly growing area of artificial intelligence
More informationWireless Networking with Selfish Agents. Li (Erran) Li Center for Networking Research Bell Labs, Lucent Technologies
Wireless Networking with Selfish Agents Li (Erran) Li Center for Networking Research Bell Labs, Lucent Technologies erranlli@dnrc.bell-labs.com Today s Wireless Internet 802.11 LAN Internet 2G/3G WAN Infrastructure
More informationThe Price of Anarchy in an Exponential Multi-Server
The Price of Anarchy in an Exponential Multi-Server Moshe Haviv Tim Roughgarden Abstract We consider a single multi-server memoryless service station. Servers have heterogeneous service rates. Arrivals
More informationHeuristic Optimization Population Control & Objectives
Heuristic Optimization Population Control & Objectives José M PEÑA (jmpena@fi.upm.es) (Universidad Politécnica de Madrid) 1 Population & Objective Population Management: Population size Micro-populations
More informationIntroduction to Artificial Intelligence. Prof. Inkyu Moon Dept. of Robotics Engineering, DGIST
Introduction to Artificial Intelligence Prof. Inkyu Moon Dept. of Robotics Engineering, DGIST Chapter 9 Evolutionary Computation Introduction Intelligence can be defined as the capability of a system to
More informationSpontaneous Cooperation under Anarchy
Spontaneous Cooperation under Anarchy 1 International Cooperation Cooperation, as you should recall, is part and parcel with our distributive games (the mixed motive games) between pure coordination and
More informationSCHOOL OF MATHEMATICS AND STATISTICS Attempt all the questions. The allocation of marks is shown in brackets.
SCHOOL OF MATHEMATICS AND STATISTICS Game Theory Autumn Semester 2017 18 2 hours and 30 minutes Attempt all the questions. The allocation of marks is shown in brackets. Please leave this exam paper on
More informationLecture #2: Introduction to Game Theory. Prof. Dr. Sven Seuken
Lecture #2: Introduction to Game Theory Prof. Dr. Sven Seuken 27.2.2012 Outline 1. Recap of last lecture 2. Go over Game Theory concepts 3. Play some in-class experiments 4. Discussion 5. Questions Quick
More informationChapter 13 Outline. Challenge: Intel and AMD s Advertising Strategies. An Overview of Game Theory. An Overview of Game Theory
Chapter 13 Game Theory A camper awakens to the growl of a hungry bear and sees his friend putting on a pair of running shoes. You can t outrun a bear, scoffs the camper. His friend coolly replies, I don
More informationAnalytic Preliminaries for Social Acceptability of Legal Norms &
Analytic Preliminaries for Social Acceptability of John Asker February 7, 2013 Yale Law School Roadmap Build on case studies of environments where extra-legal norms are at least as important as formal
More informationMultiagent Systems: Rational Decision Making and Negotiation
Introduction Multiagent Systems: Rational Decision Making and Negotiation Ulle Endriss (ue@doc.ic.ac.uk) Course website: http://www.doc.ic.ac.uk/ ue/mas-2005/ Ulle Endriss, Imperial College London 1 Introduction
More informationTerry College of Business - ECON 7950
Terry College of Business - ECON 7950 Lecture 2: Game Theory Basics Primary reference: Dixit and Skeath, Games of Strategy, Ch. 2 and 4. Game Theory A framework for understanding behavior by players who
More informationGrowing Artificial Societies
Growing Artificial Societies Social Science from the Bottom Up presentation by Ian Nunn 1 Lecture Based On Book: Growing Artificial Societies by Joshua Epstein and Robert Axtell Software: Ascape, Sugarscape
More informationBidding for Sponsored Link Advertisements at Internet
Bidding for Sponsored Link Advertisements at Internet Search Engines Benjamin Edelman Portions with Michael Ostrovsky and Michael Schwarz Industrial Organization Student Seminar September 2006 Project
More informationFAIRNESS AND SHORT RUN PRICE ADJUSTMENT IN POSTED OFFER MARKETS *
Working Paper 03-60 Economics Series 24 November 2003 Departamento de Economía Universidad Carlos III de Madrid Calle Madrid, 126 28903 Getafe (Spain) Fax (34) 91 624 98 75 FAIRNESS AND SHORT RUN PRICE
More informationIncentives in Crowdsourcing: A Game-theoretic Approach
Incentives in Crowdsourcing: A Game-theoretic Approach ARPITA GHOSH Cornell University NIPS 2013 Workshop on Crowdsourcing: Theory, Algorithms, and Applications Incentives in Crowdsourcing: A Game-theoretic
More informationMOT Seminar. John Musacchio 4/16/09
Game Theory MOT Seminar John Musacchio johnm@soe.ucsc.edu 4/16/09 1 Who am I? John Musacchio Assistant Professor in ISTM Joined January 2005 PhD from Berkeley in Electrical l Engineering i Research Interests
More informationGlobalization and Social Networks
Globalization and Social Networks Georg Durnecker University of Mannheim Fernando Vega-Redondo European University Institute September 14, 2012 Duernecker & Vega-R. (Mannheim & EUI) Globalization and Social
More informationA forgiving strategy for the Iterated Prisoner s Dilemma
Copyright JASSS Colm O Riordan (2000) A forgiving strategy for the Iterated Prisoner s Dilemma Journal of Artificial Societies and Social Simulation vol. 3, no. 4,
More informationDeterministic Crowding, Recombination And Self-Similarity
Deterministic Crowding, Recombination And Self-Similarity Bo Yuan School of Information Technology and Electrical Engineering The University of Queensland Brisbane, Queensland 4072 Australia E-mail: s4002283@student.uq.edu.au
More informationSummer 2003 (420 2)
Microeconomics 3 Andreas Ortmann, Ph.D. Summer 2003 (420 2) 240 05 117 andreas.ortmann@cerge-ei.cz http://home.cerge-ei.cz/ortmann Week of June 2, lecture 8: Game theory: Wrapping up the previous week
More informationPSO Algorithm for IPD Game
PSO Algorithm for IPD Game Xiaoyang Wang 1 *, Yibin Lin 2 1Business School, Sun-Yat Sen University, Guangzhou, Guangdong, China,2School of Software, Sun-Yat Sen University, Guangzhou, Guangdong, China
More informationBargaining with Neighbors: Is Justice Contagious?
Bargaining with Neighbors: Is Justice Contagious? Jason Alexander Brian Skyrms October 5, 1999 Abstract We investigate evolutionary dynamics for the simplest symmetric bargaining game in two different
More informationIII COST MP 0801 on Physics of Conflicts & Competition May 2011 Eurandom, Eindhoven
Concepts and interpretation: Human Resources Excellence in Human Being Behavior III COST MP 0801 on Physics of Conflicts & Competition 18-20 May 2011 Eurandom, Eindhoven Carmen Costea ASE Bucharest Physics
More informationWhat is Evolutionary Computation? Genetic Algorithms. Components of Evolutionary Computing. The Argument. When changes occur...
What is Evolutionary Computation? Genetic Algorithms Russell & Norvig, Cha. 4.3 An abstraction from the theory of biological evolution that is used to create optimization procedures or methodologies, usually
More informationEconomic Consequences of Global Accounting Convergence: An Experimental Study of a Coordination Game
Doi:10.11640/tjar.3.2013.04 Online First Version Economic Consequences of Global Accounting Convergence: An Experimental Study of a Coordination Game Satoshi Taguchi Faculty of Commerce, Doshisha University
More informationGame Theory DR. ÖZGÜR GÜRERK UNIVERSITY OF ERFURT WINTER TERM 2012/13. What can we model as a game?
Game Theory 2. Strategic Games DR. ÖZGÜR GÜRERK UNIVERSITY OF ERFURT WINTER TERM 2012/13 What can we model as a game? Example: Firm behavior Players: Firms Actions: Prices that can be set by the firms
More informationIntroduction to Microeconomic Theory
Introduction to Microeconomic Theory Lectures in Microeconomic Theory Fall 2010, Part 1 07.07.2010 G.B. Asheim, ECON4230-35, #1 1 Why Microeconomics once more? Some new subjects Game theory new to some
More informationGraham Romp Dynamic Game Theory.
Graham Romp Dynamic Game Theory. In the previous chapter we focused on static games. However for many important economic applications we need to think of the game as being played over a number of time
More informationHave We Really Lost The Space Acquisition Recipe Or Are We Simply Trying To Go Where No [One] Has Gone Before?
Have We Really Lost The Space Acquisition Recipe Or Are We Simply Trying To Go Where No [One] Has Gone Before? 2009 Flight Software Workshop Dr. Douglas J. Buettner, Space Based Surveillance Division,
More informationSymmetric Information Benchmark Begin by setting up a comparison situation, where there is no information asymmetry. Notation:
ECO 37 Economics of Uncertainty Fall Term 009 Notes for Lectures 7. Job Market Signaling In this market, the potential employee s innate skill or productive capability on the job matters to an employer
More informationAnalysis of Corporate Social Responsibility Based on Evolutionary Game Theory
International Conference on Management Engineering and Management Innovation (ICMEMI 2015) Analysis of Corporate Social Responsibility Based on Evolutionary Game Theory Xiong XIE1,a, Mei-Yan CHEN2,b, Xiao-Wen
More informationGame Theory Approach for Interactive Wind Farm Control
Game Theory Approach for Interactive Wind Farm Control Arman Kiani and Robin Franz Institute of Automatic Control Engineering, Technische Universitat Muenchen, Germany, Active Adaptive Control Laboratory,
More informationbiologically-inspired computing lecture 19 Informatics luis rocha 2015 INDIANA UNIVERSITY biologically Inspired computing
lecture 19 -inspired Sections I485/H400 course outlook Assignments: 35% Students will complete 4/5 assignments based on algorithms presented in class Lab meets in I1 (West) 109 on Lab Wednesdays Lab 0
More informationLECTURE 7: Reaching Agreements
Negotiation LECTURE 7: Reaching Agreements The process of several agents searching for an agreement e.g. about price. Reaching consensus An Introduction to MultiAgent Systems http://www.csc.liv.ac.uk/~mjw/pubs/imas
More informationModeling of competition in revenue management Petr Fiala 1
Modeling of competition in revenue management Petr Fiala 1 Abstract. Revenue management (RM) is the art and science of predicting consumer behavior and optimizing price and product availability to maximize
More informationChapter 15 Oligopoly
Goldwasser AP Microeconomics Chapter 15 Oligopoly BEFORE YOU READ THE CHAPTER Summary This chapter explores oligopoly, a market structure characterized by a few firms producing a product that mayor may
More informationECONOMICS 103. Topic 2: Specialization & Trade
ECONOMICS 103 Topic 2: Specialization & Trade Key concepts: production possibilities, absolute advantage, comparative advantage, trade, gains from trade, economic efficiency. Model: production possibility
More informationarxiv: v1 [econ.em] 12 Jul 2018
Analysis of a Dynamic Voluntary Mechanism Public Good Game Dmytro Bogatov Worcester Polytechnic Institute, Worcester, MA 169 USA (e-mail: dmytro@dbogatov.org) arxiv:187.4621v1 [econ.em] 12 Jul 218 Abstract:
More informationA Negotiation Meta Strategy Combining Trade-off and Concession Moves. Team GiT GuD Ashley Dyer, Aaron Post, Paul Quint
A Negotiation Meta Strategy Combining Trade-off and Concession Moves Team GiT GuD Ashley Dyer, Aaron Post, Paul Quint Automated Negotiation Main key for autonomous agent interaction Negotiation Protocols:
More informationTotal /18 /18 /16 /15 /15 /18 /100. Economics 142 Final Exam NAME Vincent Crawford Winter 2008
2 3 4 5 6 Total /8 /8 /6 /5 /5 /8 / Economics 42 Final Exam NAME Vincent Crawford Winter 28 Your grade from this exam is two thirds of your course grade. The exam ends promptly at 2:3, so you have three
More informationNote on webpage about sequential ascending auctions
Econ 805 Advanced Micro Theory I Dan Quint Fall 2007 Lecture 20 Nov 13 2007 Second problem set due next Tuesday SCHEDULING STUDENT PRESENTATIONS Note on webpage about sequential ascending auctions Everything
More informationExam #2 (100 Points Total) Answer Key
Exam #2 (100 Points Total) Answer Key 1. A Pareto efficient outcome may not be good, but a Pareto inefficient outcome is in some meaningful sense bad. (a) (5 points) Give an example or otherwise explain,
More informationAnalyze different types of non zero sum games. Hawk vs. Dove game. Advertising problem. Problem of companies polluting the environment.
Unit 3 GME THEORY Lesson 31 Learning Objective: nalyze different types of non zero sum games Hawk vs. game. dvertising problem. Problem of companies polluting the environment. Rationalizability or Iterative
More informationApproximation in Algorithmic Game Theory
Approximation in Algorithmic Game Theory Robust Approximation Bounds for Equilibria and Auctions Tim Roughgarden Stanford University 1 Motivation Clearly: many modern applications in CS involve autonomous,
More informationLecture 2: March Introduction. Computational Game Theory Spring Semester, 2011/2012
Computational Game Theory Spring Semester, 2011/2012 Lecture 2: March 14 Lecturer: Amos Fiat Scribe: Hadas Zur & Alon Ardenboim 2.1 Introduction Informally, a set of strategies is a Nash equilibrium if
More informationLecture 13(ii) Announcements. 1. Application of Game Theory: Duopoly (The one-shot game case)
Lecture 13(ii) Announcements Check out Final Exam OneStop Page at the very bottom of Moodle One stop shopping for all your final preparation needs, including questions from previous finals. 1. Application
More informationAGENT-BASED MODELLING OF A SOCIAL DILEMMA IN MODE CHOICE BASED ON TRAVELERS EXPECTATIONS AND SOCIAL LEARNING MECHANISMS
Advanced OR and AI Methods in Transportation AGENT-BASED MODELLING OF A SOCIAL DILEMMA IN MODE CHOICE BASED ON TRAVELERS EXPECTATIONS AND SOCIAL LEARNING MECHANISMS Yos SUNITIYOSO 1, Shoji MATSUMOTO 2
More informationIF YOU ARE SO RICH, WHY AREN T YOU SMART? Nobuyuki Hanaki Juliette Rouchier
Proceedings of the 2013 Winter Simulation Conference R. Pasupathy, S.-H. Kim, A. Tolk, R. Hill, and M. E. Kuhl, eds. IF YOU ARE SO RICH, WHY AREN T YOU SMART? Nobuyuki Hanaki Juliette Rouchier Aix-Marseille
More informationThe Success and Failure of Tag-Mediated Evolution of Cooperation
The Success and Failure of Tag-Mediated Evolution of Cooperation Austin McDonald and Sandip Sen Mathematics and Computer Science Department University of Tulsa {austin, sandip}@utulsa.edu Abstract. Use
More information