Top Banner
Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)
35

Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Mar 27, 2015

Download

Documents

Logan MacKenzie
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Pushing the Envelope:new research topics at the interface of cs and econ/gt

Yoav Shoham

Stanford University

(many debts are due)

Page 2: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 2

Primary areas of interaction so far

• Computing solution concepts, primarily NE

• Multi-agent learning

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

• Mechanism design, in particular auctions

Page 3: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 3

Talk Outline

• Computing solution concepts, primarily NE

– The role of NE unclear

• Multi-agent learning

– Ditto

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

– Other forms of compactness, and what about coalitional games?

• Mechanism design, in particular auctions

– Behavioral Mechanism design

• Beyond GT: Algorithmic Institutional Design

Page 4: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 4

A game with a trivial, unique NE

Heads Tails

Heads 1,-1 -1,1

Tails -1,1 1,-1

Rock Paper Scissors

Rock 0,0 -1,1 1,-1

Paper 1,-1 0,0 -1,1

Scissors -1,1 1,-1 0,0

Matching Pennies Rochambeau (Rock-Paper-Scissors)

Page 5: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 5

A game with a trivial, unique NE

Heads Tails

Heads 1,-1 -1,1

Tails -1,1 1,-1

Rock Paper Scissors

Rock 0,0 -1,1 1,-1

Paper 1,-1 0,0 -1,1

Scissors -1,1 1,-1 0,0

Matching Pennies Rochambeau (Rock-Paper-Scissors)

(www.worldrps.com)

Page 6: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 6

A game with a trivial, unique NE

Heads Tails

Heads 1,-1 -1,1

Tails -1,1 1,-1

Rock Paper Scissors

Rock 0,0 -1,1 1,-1

Paper 1,-1 0,0 -1,1

Scissors -1,1 1,-1 0,0

Matching Pennies Rochambeau (Rock-Paper-Scissors)

(www.worldrps.com)Lesson: Nash equilibrium not necessarily instructive

Page 7: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 7

Some Intuition about Learning

Left Right

Up 1,0 3,2

Down 2,1 4,0

Stackelberg Game

Page 8: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 8

Some Intuition about Learning

Left Right

Up 1,0 3,2

Down 2,1 4,0

Stackelberg Game

Lesson: can’t separate learning from teaching

Page 9: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 10

Five Distinct Research Agendas in MAL

• Computation: Quick-and-dirty method for (e.g.) NE

• Social science: How people (institutions, animals…) learn.

• Game theory puritanism: Equilibria of learning strategies.

• Distributed control: Learning in common-payoff games.

• Targeted learning: Learning when you have some sense of how your opponents might behave.

Page 10: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 11

Lesson: Need to take NE with a grain of salt

• Beautiful, clever

• Makes it hard to back off from assumptions of perfect rationality; can we have an alternative, “constructive” game theory?

• In any event, “best response” computation merits as much attention as eqm

Page 11: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 12

Talk Outline

• Computing solution concepts, primarily NE

– The role of NE unclear

• Multi-agent learning

– Ditto

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

– Other forms of compactness, and what about coalitional games?

• Mechanism design, in particular auctions

– Behavioral Mechanism design

• Beyond GT: Algorithmic Institutional Design

Page 12: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 13

On compact representations

• Compact representations are fine; need more– Programming constructs in strategy descriptions (“programmatic

rationality”)– Partial games (e.g., logic-based game description)

• What about coalitional games?

Page 13: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 14

Marginal Contribution Nets

• Games represented by sets of rules

pattern value

{ a & b & c } 5

• Value of a group S equals the sum of the values of the rules S satisfies

v(S) = r : S satisfies r} v(r)

• Focus on conjunction & negation in pattern

Page 14: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 15

Conciseness of MC-Nets

Theorem MC-Nets generalize the multi-issue representation of [CS04]

Theorem MC-Nets generalize the graphical representation of [DP94]

Page 15: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 16

Computational Leverage

• Shapley value can be efficiently computed in MC-nets

– Exploiting Additivity and Symmetry

• Determining membership in core is hard, but one can determine membership in time exponential in treewidth

– Determining emptiness, or finding an arbitrary member of a non-empty core, are no harder

Page 16: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 17

Talk Outline

• Computing solution concepts, primarily NE

– The role of NE unclear

• Multi-agent learning

– Ditto

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

– Other forms of compactness, and what about coalitional games?

• Mechanism design, in particular auctions

– Behavioral Mechanism design

• Beyond GT: Algorithmic Institutional Design

Page 17: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 18

Recall some results from auction theory

• Informal observations– Dutch = First-price, sealed bid– English Second-price, sealed bid (cf. proxy bidding)– Japanese ≠ English– Second-price and Japanese have dominant strategies

• For precise analyses, need to distinguish between– Common values and independent values (winner’s curse)– Risk averse, risk-neutral and risk-seeking bidders

• Formal results speak to:– Whether an auction is “incentive compatible”– Whether the auction is “efficient”– Whether the auction is “revenue maximizing”

Page 18: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 19

Example of BMD: Online marketing

• The X5 story

• What are we optimizing for?

• Behavioral requirements (BMD) (ack: Moshe Tennenholtz)

– # sign-ups

– # return visits (magic number: 5)

– Message injection

– Product education

– Truthful consumer surveys

• Yields a new perspective on existing mechanisms

• Suggests new mechanisms

Page 19: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 20

Some new truths about auctions, from the perspective of marketing

• First-price sealed-bid auction ≠ Dutch auction

• Second-price sealed-bid auction ≠ English auction

• Dominant-strategy mechanisms can be suboptimal

• Barter- and multiple-currency markets might trump markets with universal currency

Page 20: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 21

Some new, marketing-oriented mechanisms

• Tournament auction– Infinitely many equilibria

• Average-price auction– Giving the little guy a chance

• Team bidding– Cooperation

• Community auction– Coopetition

• Online collectibles– The marketing advantages of barter systems

• Preference auction– Win-win for the auctioneer and buyers

Page 21: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 22

Tournament auction

A series of sealed-bid auctions; X% make it to the next day; person with highest remaining points wins.

Page 22: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 23

Tournament auction

Other activities added to basic tournament auction

Page 23: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 24

Inserting a population game into an auction

Capturing information about consumers and their views of others; the latter is particularly truthful.

Page 24: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 25

Average Price Game

The consumer who bids closest to the average of all bids wins the

prize.

Page 25: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 26

Team Bidding

Bidders form teams and pool their bids.

Page 26: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 27

… Cariocas’ Community Auction

A “global bid” triggers the close

of multiple auctions.

Community Auction

Page 27: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 28

Online collectibles

Online collection of digital objects, initially assembled by various

online activities.

Page 28: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 29

Online collectibles

… and then exchanged via online

barter

Page 29: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 30

Main takeaways

• Marketing considerations completely change the rules of the game. Some lessons of BMD:

– new design criteria

– new perspectives on existing mechanisms

– new mechanisms

• Many applications beyond marketing. Example: Captchas, ESP

• A lot more work is needed before this becomes a science

Page 30: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 31

Talk Outline

• Computing solution concepts, primarily NE

– The role of NE unclear

• Multi-agent learning

– Ditto

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

– Other forms of compactness, and what about coalitional games?

• Mechanism design, in particular auctions

– Behavioral Mechanism design

• Beyond GT: Algorithmic Institutional Design

Page 31: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 32

Algorithmic Institutional Design (ack: Mike Munie)

• What is better: The EE or CS qual structure at Stanford?

• Similar for job interviews, admissions, consumer surveys, etc

• Reminiscent of, but distinct from, the “secretary problem”

• The answer: Depends on what you’re optimizing for. And even given that, depends.

Page 32: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 33

Formal Model, continued

Page 33: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 34

Results

• Multiple versions– Single prof?– Single student?– Parallel or sequential?

• Sample results– Even in simplest case, selecting an optimal set of questions is NP-

Hard, and is not submodular, so there is a not an obvious approximation algorithm

– Sequentiality can be maximally helpful– In the multiagent setting, even deciding between committee

structures is NP-Hard– *Seems* like there are well behaved special cases

Page 34: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 35

Talk Outline

• Computing solution concepts, primarily NE

– The role of NE unclear

• Multi-agent learning

– Ditto

• Compact games (graphical games, MAIDs, game networks, local-effect games, social networks, …)

– Other forms of compactness, and what about coalitional games?

• Mechanism design, in particular auctions

– Behavioral Mechanism design

• Beyond GT: Algorithmic Institutional Design

Page 35: Pushing the Envelope: new research topics at the interface of cs and econ/gt Yoav Shoham Stanford University (many debts are due)

Stanford, April 2007 BAGT Symposium 36

thank you!