Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Optimisation of short memory strategies in the ...
Search
Nikoleta
June 04, 2017
Science
0
55
Optimisation of short memory strategies in the Iterated Prisoners Dilemma
Wales Mathematics Colloquium 2017.
Nikoleta
June 04, 2017
Tweet
Share
More Decks by Nikoleta
See All by Nikoleta
A trip to earth science with python as a companion
nikoletav3
0
47
Arcas: Using Python to access open research literature
nikoletav3
1
180
Testing Research Software
nikoletav3
0
310
Arcas
nikoletav3
0
460
SSI Selection Day
nikoletav3
0
400
SWORDS-03-10-2016
nikoletav3
0
51
PyCon UK 2016
nikoletav3
0
160
Other Decks in Science
See All in Science
07_浮世満理子_アイディア高等学院学院長_一般社団法人全国心理業連合会代表理事_紹介資料.pdf
sip3ristex
0
650
06_浅井雄一郎_株式会社浅井農園代表取締役社長_紹介資料.pdf
sip3ristex
0
670
Cross-Media Technologies, Information Science and Human-Information Interaction
signer
PRO
3
31k
機械学習 - SVM
trycycle
PRO
1
910
データマイニング - グラフ構造の諸指標
trycycle
PRO
0
200
データベース15: ビッグデータ時代のデータベース
trycycle
PRO
0
370
Lean4による汎化誤差評価の形式化
milano0017
1
340
DMMにおけるABテスト検証設計の工夫
xc6da
1
1.1k
NASの容量不足のお悩み解決!災害対策も兼ねた「Wasabi Cloud NAS」はここがスゴイ
climbteam
1
180
Quelles valorisations des logiciels vers le monde socio-économique dans un contexte de Science Ouverte ?
bluehats
1
550
2025-06-11-ai_belgium
sofievl
1
170
Gemini Prompt Engineering: Practical Techniques for Tangible AI Outcomes
mfonobong
2
180
Featured
See All Featured
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
12
1.2k
Balancing Empowerment & Direction
lara
5
700
The Straight Up "How To Draw Better" Workshop
denniskardys
238
140k
Embracing the Ebb and Flow
colly
88
4.9k
The Art of Programming - Codeland 2020
erikaheidi
56
14k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
230
22k
Context Engineering - Making Every Token Count
addyosmani
7
280
Designing Experiences People Love
moore
142
24k
[Rails World 2023 - Day 1 Closing Keynote] - The Magic of Rails
eileencodes
37
2.6k
Leading Effective Engineering Teams in the AI Era
addyosmani
7
560
Side Projects
sachag
455
43k
Thoughts on Productivity
jonyablonski
70
4.9k
Transcript
Optimisation of short memory strategies in the Iterated Prisoners Dilemma
Nikoleta E. Glynatsi Supervised by: Dr. Vincent Knight Dr. Jonathan Gillard
(3, 3) (0, 5) (5, 0) (1, 1)
(3, 3) (0, 5) (5, 0) (1, 1) (R, P,
S, T) = (3, 1, 0, 5)
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995
2000 2005 2010 2015 0 20 40 60 80 100 number of records Articles per Year (N=1145)
CC CD DC DD C D C D C D
C D p1 1 − p1 p2 1 − p2 p3 1 − p3 p4 1 − p4 p = (p1 , p2 , p3 , p4 ) ∈ R4 [0,1]
Christopher Lee, Marc Harper, and Dashiell Fryer. The art of
war: Beyond memory-one strategies in population games. 2015.
How good are memory one strategies ?
CC CD DC DD
M = p1 q1 p1 (−q1
+ 1) q1 (−p1 + 1) (−p1 + 1)(−q1 + 1) p2 q3 p2 (−q3 + 1) q3 (−p2 + 1) (−p2 + 1)(−q3 + 1) p3 q2 p3 (−q2 + 1) q2 (−p3 + 1) (−p3 + 1)(−q2 + 1) p4 q4 p4 (−q4 + 1) q4 (−p4 + 1) (−p4 + 1)(−q4 + 1)
maxp uq (p) such that p ∈ R4 [0,1]
Lemma uq(p) = 1 2 pQpT + cT p +
a 1 2 p ¯ QpT + ¯ cT p + ¯ a Q, ¯ Q ∈ R4×4 c, ¯ c ∈ R4×1 a, ¯ a ∈ R
maxp uq (p) such that p ∈ R4 [0,1]
maxp uq (p) such that p ∈ R4 [0,1] subject
to p1 = p2 = p3 = p4 = p
Lemma uq(p) = n2p2 + n1p + n0 d1p +
d0 n2 = −(q1 − q2 − 2q3 + 2q4) n1 = −q1 + 2q2 + 5q3 − 7q4 − 1 n0 = q2 − 5q4 − 1 d1 = q1 − q2 − q3 + q4 d0 = q2 − q4 − 1
q = 1, 1, 0, 2 3 0 1 p
0 1 2 3 4 5 theoretic simulated
q = 1, 1, 0, 2 3 0 1 p
0 1 2 3 4 5 theoretic simulated uq (p) = −4p2 3 + 14p 3 − 10 3 2p 3 − 2 3
q = 1, 1, 0, 2 3 0 1 p
0 1 2 3 4 5 theoretic simulated uq (p) = −4p2 3 + 14p 3 − 10 3 2p 3 − 2 3 = −2p + 5
q = 1, 0, 1, 1 3 0 1 p
0 1 2 3 4 5 theoretic simulated
q = 1, 0, 1, 1 3 0 1 p
0 1 2 3 4 5 theoretic simulated uq (p) = p2 3 + 8p 3 − 10 3 p 3 − 4 3
q = 1, 0, 1, 1 3 0 1 p
0 1 2 3 4 5 theoretic simulated uq (p) = p2 3 + 8p 3 − 10 3 p 3 − 4 3 = p + 2
q = 2 3 , 0, 2 3 , 1
3 0 1 p 0 1 2 3 4 5 theoretic simulated
q = 2 3 , 0, 2 3 , 1
3 0 1 p 0 1 2 3 4 5 theoretic simulated uq (p) = 2p 3 − 8 3 p 3 − 4 3
q = 2 3 , 0, 2 3 , 1
3 0 1 p 0 1 2 3 4 5 theoretic simulated uq (p) = 2p 3 − 8 3 p 3 − 4 3 = 2
q = 2 3 , 1 3 , 1 3
, 0 0 1 p 0 1 2 3 4 5 theoretic simulated
q = 2 3 , 1 3 , 1 3
, 0 0 1 p 0 1 2 3 4 5 theoretic simulated uq (p) = p2 3 − 2p 3 − 2 3 −2 3
q = 2 3 , 1 3 , 1 3
, 0 0 1 p 0 1 2 3 4 5 theoretic simulated uq (p) = p2 3 − 2p 3 − 2 3 −2 3 = − p2 2 + p + 1
Lemma (Indifferent) −q1 + q2 + 2q3 − 2q4 =
0 and (q2 − q4 − 1)(q1 − 2q2 − 5q3 + 7q4 + 1) − (q2 − 5q4 − 1)(q1 − q2 − q3 + q4 ) = 0. Proof. uq (p) = n2 p2 + n1 p + n0 d1 p + d0 = a0 n2 p2 + n1 p + n0 = a0 d1 p + a0 d0 n2 = 0 n1 d0 = d1 n0
Lemma (Linear) (q1 q4 − q2 q3 + q3 −
q4 )(4q1 − 3q2 − 4q3 + 3q4 − 1) = 0 Proof. uq (p) = n2 p2 + n1 p + n0 d1 p + d0 = a1 p + a0 n2 p2 + n1 p + n0 = a1 d1 p2 + (d1 a0 + a1 d0 )p + a0 d0 n2 = d1 a1 n1 d0 = d1 n0 + a1 d0
Lemma (Quadratic) (q1 − q2 − q3 + q4 )
= 0, (q1 q4 − q2 q3 + q3 − q4 )(4q1 − 3q2 − 4q3 + 3q4 − 1) = 0 and q2 − q4 − 1 = 0 Proof. uq (p) = n2 p2 + n1 p + n0 d1 p + d0 = a2 p2 + a1 p + a0 n2 p2 + n1 p + n0 = d1 a2 p3 + (a1 d1 + d0 a2 )p2 + (d1 a0 + a1 d0 )p + a0 d0 a1 d1 = 0 n2 = d1 a1 + d0 an2 n1 d0 = d1 n0 + a1 d0
du dp = m2 p2 + m1 p + m0
(d1 p + d0 )2 p uq p− p+ p uq p− p+ p uq p− p+ p uq p− p+
Theorem (Optimization of purely random player) Sq = 0, p±
, 1 0 < p± < 1, p± = −d0 d1 p∗ = argmax p∈Sq uq (p)
q = 7 8 , 7 16 , 3 8
, 0 0.0 0.2 0.4 0.6 0.8 1.0 p 0 1 2 3 4 5 theoretical p* simulated
q = 1 3 , 2 3 , 1, 0
0.0 0.2 0.4 0.6 0.8 1.0 p 0 1 2 3 4 5 theoretical p* simulated
q(1), q(2), q(3) . . . q(N) max p 1
N N i=1 uq (i)(p)
q(1), q(2), q(3) . . . q(N) max p 1
N N i=1 uq (i)(p) max p u 1 N N i=1 q(i) (p)
0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3
4 5 Tournament size N=9 q u p* simulated
p∗ = argmaxS q(1),...,q(n) u(p) where, | Sq(1),...,q(n) |≤ 2N
+ 2
p∗ = argmaxS q(1),...,q(n) u(p) where, | Sq(1),...,q(n) |≤ 2N
+ 2 @NikoletaGlyn https://github.com/Nikoleta-v3