CZ3005

Artificial Intelligence

Game Theory
Assoc Prof Bo AN

www.ntu.edu.sg/home/boan
Email: [email protected]
Office: N4-02b-55

1
 What is Game Theory?

• Game theory studies settings where multiple parties (agents)

each have
– different preferences (utility functions),
– different actions that they can take
• Each agent’s utility (potentially) depends on all agents’ actions
– What is optimal for one agent depends on what other agents do
– Very circular!
• Game theory studies how agents can rationally form beliefs over
what other agents will do, and (hence) how agents should act
– Useful for acting as well as predicting behavior of others

• John von Neumann

2
What is Game Theory?

Economics Politics

Games Biology
3
Normal Form Games – An Example

• List of players, strategies, payoffs

• Simultaneous
• Zero-sum here but not necessary Player B
• Analysis:
– What should they do? Rock
– Advice to player A

0, 0 1,-1 -1,1

Player A Rock
-1, 1 0, 0 1, -1

1, -1 -1, 1 0, 0

4
Nash Equilibrium

• Each agent is selfish

• Each agent makes decision based on what he thinks others
would do
• No one can do better by changing strategy solely

5
Nash Equilibrium

Player B

Paper 1/3 Rock 1/3 Scissors 1/3

Paper
0, 0 1,-1 -1,1
1/3
Player A Rock -1, 1 0, 0 1, -1
1/3
1, -1 -1, 1 0, 0
Scissors

1/3
6
 Nash Equilibrium

Nash equilibrium

confess dominates silent

Cooperation would be better for both!

But, rational for both to defect!

7
Nash Equilibrium

• In general, we will say that two strategies s1 (for i)

and s2 (for j) are in Nash equilibrium if:
1. under the assumption that agent i plays s1, agent j can do no better than
play s2; and
2. under the assumption that agent j plays s2, agent i can do no better than
play s1.
• Neither agent has any incentive to deviate from a
Nash equilibrium
• Unfortunately:
1. Not every interaction scenario has
a pure strategy Nash equilibrium
2. Some interaction scenarios have more
than one pure strategy Nash equilibrium

8
Matching Pennies

• Players i and j simultaneously choose the face of a

coin, either “heads” or “tails”.
• If they show the same face, then i wins, while if they
show different faces, then j wins.
• The Payoff Matrix:

9
Mixed Strategies for Matching Pennies

• NO pair of strategies forms a pure strategy NE:

whatever pair of strategies is chosen, somebody will
wish they had done something else.
• The solution is to allow mixed strategies:
– play “heads” with probability 0.5
– play “tails” with probability 0.5.
• This is a NE strategy.

10
Games: Complete Information
v.s. Incomplete Information

Games with
complete information

Chess Go

Games with
incomplete information

Negotiation Poker
11
Games: Normal Form v.s. Sequence
Form

12
Texas Hold’em Poker

• Preflop: two private cards are dealt to each player, followed by a

betting round; players can either check, bet or fold
• Flop: three public cards are dealt, followed by a betting round
• Turn: a fourth public card is dealt, followed by a betting round
• River: a last public card is dealt, followed by a betting round
• The game ends when
– Only one player is left, all the other players fold
– A showdown; a hand with the best 5 cards using both the two private cards
and the five public cards wins

13
Architecture of Libratus

Original game Abstracted

game
• action abstraction Automated abstraction
Abstraction • card abstraction
(offline) • took the game size Compute
from 10161 to 1012 Nash
Reverse model
Nash equilibrium Nash equilibrium
Equilibrium • CFR
Finding • CFR+
(offline) • Monte Carlo CFR

Decomposition
• endgame solving
and Subgame
• subgame re-solving
Refinement
• max-margin subgame refinement
(online)

14
Global Challenges for Security

Key challenges: Limited resources, surveillance

15
Stackelberg Games
Randomization: Increase Cost and Uncertainty to Attackers

• Security allocation
– Target weights
– Opponent reaction
• Stackelberg: Security forces commit first
• Optimal allocation: Weighted random
– Strong Stackelberg Equilibrium
Attacker

Target #1 Target #2
Defender Target #1 4, -3 -1, 1
Target #2 -5, 5 2, -1
16
Game Theory for Security: Applications
Game Theory + Optimization + Uncertainty + Learning + …
Infrastructure Security Games Green Security Opportunistic
Games Crime Games

Coast Guard Coast Guard: Ferry

Coast Guard

LAX TSA
Cyber Security
Games

Panthera/WWF
LA Sheriff USC

Argentina Airport Chile Border

India

17
 IRIS: Federal Air Marshals Service [2009]
Scale Up Number of Defender Strategies
Strategy 1 Strategy 2 Strategy 3 Strategy 1 Strategy 2 Strategy 3

Strategy Str
1 ate
gy
1

Strategy
2 Str
ate
gy
2
Strategy
3
Str
ate
gy
Strategy 3
4

Str
ate
Strategy gy
5 4

Str
Strategy
ate
6
gy
5

Str
ate
gy
6

• 1000 Flights, 20 air marshals: 1041 combinations

– ARMOR out of memory

• Not enumerate all combinations combinations

– Branch and price
o Branch & bound + column generation

18
PROTECT: Randomized Patrol Scheduling [2011]
Coordination (Scale-up) and Ferries (Continuous Space/time)
t2 t3

t7
t4
t1

t6 t5

19
 PAWS: Protection Assistant for Wildlife
Security Trials in Uganda and Malaysia [2014]
• Important lesson: Geography!

Uganda Andrew Lemieux Malaysia Panthera

Malaysia

20
PAWS Deployed in 2015 in Southeast
Asia (with Panthera and WWF)
PAWS Version 2: Features
• Street map
– Ridgelines, rivers/streams

• Species Distribution Models (SDMs)

– From data points to distribution map

Indonesia Malaysia

21