- PDF Free Download

Size: px

Start display at page:

Download ""

みそらはやしもと
5 years ago
Views:

1 Jupiter User Guide

3 Jupiter Genius[2] Genius Jupiter Jupiter Stacked Alternating Offers Protocol(SAOP)[1] Jupiter 1 Genius Jupiter 1 Jupiter 2 Jupiter 3 1

4 Jupiter Jupiter Jupiter Jupiter Jupiter Jupiter

5 20 3

6 Jupiter

7 1 Jupiter 1.1 Jupiter [3] 1.2 Jupiter Stacked Alternating Offers Protocol(SAOP)[1] SAOP 1 2 Offer Accept EndNegotiation Jupiter

8 Jupiter 1. EndNegotiation EndNegotiation EndNegotiation Offer Accept EndNegotiation Offer 1 Offer Offer Offer 2. Accept Offer Offer Accept 1.3 6

9 EndNegotiation EndNegotiation (bid) s n I i i (bid) (1.1) s = {(s 1, s 2,..., s n ) s i I i, i = 1, 2,..., n} (1.1) (1.1) 3 1. U(s k ; w) (1.2) (1.3) (1.4) ) U(s; w) = n w i eval(s i ) (1.2) i=1 n w i = 1 (1.3) i=1 eval(s i ) = value(s i ) arg max value(s (1.4) s i I i i w i i 0 w i 1(1 i n) value(s i ) I i s i 7

10 Jupiter w i (1 i n) value(s i )(s i I i ) (1.4) (1.2) (1.3) (1.5) 0 eval(s i ) 1 (1.5) 0 U(s; w) 1( s) 2. EndNegotiation r e (1.6) U(e) = r (1.6) 3. U d d t(0 t 1) U d (s, t; d) = U(s) d t (1.7) U d (e, t; d) = U(e) d t (1.8) 0.2, 0.4,..., 1.0 d t 1.1 8

11 : U(s) 1.2 n = 2 I 1 = 2 I 2 = 2 w 1 = 0.3 w 2 = 0.7 value(s 11 ) = 3 value(s 12 ) = 5 value(s 21 ) = 3 value(s 22 ) = 6 S S = {(s 11, s 21 ), (s 11, s 22 ), (s 12, s 21 ), (s 12, s 22 )} (s 11, s 21 ) value(s 11 ) U((s 11, s 21 ); w) = w 1 arg max value(s 1 ) + w value(s 21 ) 2 arg max value(s 2 ) s 1 I 1 s 2 I 2 = =

12 Jupiter 1.2: 2 2 U((s 11, s 22 ); w) = 0.88 U((s 12, s 21 ); w) = 0.65 U((s 12, s 22 ); w) =

13 2 Jupiter 2.1 Jupiter Jupiter Python3 python 3.5 numpy pandas py4j matplotlib Cython Cython C mac linux windows MinGW gcc 11

14 Jupiter macos OS macos Sierra Python Python3 cycler== matplotlib==2.1.1 numpy== pandas== py4j== pyparsing==2.2.0 python-dateutil==2.6.1 pytz== six== Cython=0.28 Windows OS Windows 10 Python

15 2.3 Jupiter Python3 cycler== Cython=0.28 kiwisolver==1.0.1 matplotlib==2.2.0 numpy== pandas== py4j== pyparsing==2.2.0 python-dateutil==2.7.0 pytz== six== Jupiter Jupiter 1. pip install jupiter-negotiation 2. jupiter test jupiter Jupiter Jupiter 13

16 Jupiter 2.1: Jupiter Listing 2.1: Jupiter 1 import os 2 import j u p i t e r 3 from j u p i t e r. s i m u l a t o r. j u p i t e r import J u p i t e r 4 from j u p i t e r. s i m u l a t o r. n e g o t i a t i o n R u l e import TypeOfNegotiation 5 from j u p i t e r. agents import l i n e a r A g e n t 6 from j u p i t e r. agents import concederagent 7 8 i f name == m a i n : 9 j u p i t e r p a t h = os. path. abspath ( j u p i t e r. p a t h [ 1]) 10 domain path = os. path. j o i n ( j u p i t e r p a t h, j u p i t e r /domain/ ) 11 j u p i t e r = J u p i t e r ( TypeOfNegotiation. Turn, , 13 domain path + Atlas3 / t r i a n g u l a r F i g h t. xml, 14 domain path + Atlas3 / t r i a n g u l a r F i g h t u t i l 1. xml, 15 domain path + Atlas3 / t r i a n g u l a r F i g h t u t i l 2. xml ) 16 j u p i t e r. s e t a g e n t ( linearagent, LinearAgent ) 17 j u p i t e r. s e t a g e n t ( concederagent, ConcederAgent ) 18 j u p i t e r. d o n e g o t i a t i o n ( i s p r i n t i n g=true, p r i n t t i m e s =1) 19 j u p i t e r. d i s p l a y. show ( ) Jupiter Listing main.py 2. main.py Listing main.py 14

17 2.4 Jupiter main.py Jupiter 1. jupiter Jupiter negotiationrule TypeOfNegotiation import 2. Jupiter Jupiter 4 3. Listing2.1 LinearAgent ConcederAgent str Jupiter jupiter.py Jupiter Genius Genius Jupiter str Jupiter 1. myagent.py main.py 15

18 Jupiter 2. myagent.py Listing main.py 17 jupiter.set agent(myagent, MyAgent ) Listing 2.2: myagent.py 1 import s y s 2 import os 3 from j u p i t e r. s i m u l a t o r import abstractagent 4 from j u p i t e r. s i m u l a t o r import agentaction 5 from j u p i t e r. s i m u l a t o r import a b s t r a c t U t i l i t y S p a c e 6 from j u p i t e r. s i m u l a t o r import n e g o t i a t i o n R u l e c l a s s MyAgent ( abstractagent. AbstractAgent ) : 10 d e f i n i t ( s e l f, 11 u t i l i t y s p a c e : a b s t r a c t U t i l i t y S p a c e. A b s t r a c t U t i l i t y S p a c e, 12 n e g o t i a t i o n r u l e : n e g o t i a t i o n R u l e. NegotiationRule, 13 a g e n t i d : int, 14 agent num : i n t ) : 15 s e l f. u t i l i t y s p a c e = u t i l i t y s p a c e 16 s e l f. r u l e = n e g o t i a t i o n r u l e 17 s e l f. a g e n t i d = a g e n t i d 18 s e l f. o p p o n e n t b i d = None d e f r e c e i v e a c t i o n ( s e l f, a g e n taction : agentaction. AbstractAction ) : 21 i f i s i n s t a n c e ( agentaction, agentaction. O f f e r ) : 22 s e l f. o p p o n e n t b i d = agentaction. g e t b i d ( ) d e f s e n d a c t i o n ( s e l f ) : 25 d e f g e t c o n s s e t i o n v a l u e ( ) : 26 r e t u r n ( 1. 0 s e l f. r u l e. get time now ( ) ) i f s e l f. o p p o n e n t b i d i s not None and \ 29 g e t c o n s s e t i o n v a l u e ( ) < \ 30 s e l f. u t i l i t y s p a c e. g e t u t i l i t y ( s e l f. o p p o n e n t b i d ) : 31 r e t u r n agentaction. Accept ( s e l f. a g e n t i d ) t h r e s h o l d = g e t c o n s s e t i o n v a l u e ( ) 34 b i d o f f e r = \ 35 s e l f. u t i l i t y s p a c e. g e t b i d a b o v e c o n c e s s i o n v a l u e ( t h r e s h o l d ) 36 r e t u r n agentaction. O f f e r ( s e l f. a g e n t i d, b i d o f f e r ) d e f r e c e i v e s t a r t n e g o t i a t i o n ( s e l f ) : 39 s e l f. o p p o n e n t b i d = None d e f get name ( s e l f ) : 42 r e t u r n MyAgent 16

19 2.5 Jupiter 2.5 Jupiter Jupiter Linear Conceder Boulware (Accept) (Offer) (EndNegotiation) 1 Acceptance Strategy Acceptance Strategy 2 T (t) = u max (1.0 t α ) (2.1) (2.1) T (t) u max Genius Jupiter u max = 1 (2.1) α = 1 Linear Linear Conceder (2.1) α = 0.5 Boulware (2.1) α = ANAC jupiter pip show jupiter-negotiation 17

20 Jupiter Jupiter Jupiter Notebook start.html 18

21 3 User Guide Jupiter 19

22 [1] Aydoğan, Reyhan, et al.: Alternating offers protocols for multilateral negotiation. Modern Approaches to Agent-based Complex Automated Negotiation. Springer International Publishing, [2] Lin, Raz, et al.: Genius: An integrated environment for supporting the design of generic automated negotiators. Computational Intelligence 30.1 (2014): [3] T. Baarslag, R. Aydogan, K. V. Hindriks, K. Fuijita, T. Ito, and C. M. Jonker.: The automated negotiating agents competition AI Magazine,

情報処理学会研究報告 IPSJ SIG Technical Report Vol.2015-ICS-179 No /3/20 1,a) 1,b) Bilateral Multi-issue Closed Bargaining Problem BMCBP BMCBP ANAC A Neg

情報処理学会研究報告 IPSJ SIG Technical Report Vol.2015-ICS-179 No /3/20 1,a) 1,b) Bilateral Multi-issue Closed Bargaining Problem BMCBP BMCBP ANAC A Neg 1,a) 1,b) Bilateral Multi-issue Closed Bargaining Problem BMCBP BMCBP ANAC A Negotiation Strategy based on Evolutionary Stable Strategy in Bilateral Closed Bargaining Problem Akiyuki Mori 1,a) Takayuki