Presentation 2002/1/3
Co-evolutions through Imitation Learning
Yukikazu MURAKAMI, Hiroshi SATO, Akira NAMATAME,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) It is an interesting question to answer how the society gropes its way towards efficient equilibrium in an imperfect world when self-interested agents learn each other in order to improve the rules of interaction. In this paper, we focus on the co-evolution of meta-rules of strategic interactions in the negotiation situations. We formulate negotiation situations as hawk-dove games. It is known the mixed strategy will result in an efficient equilibrium in hawk-dove games. In this paper, we introduce co-evolutionary dynamics with local matching and investigate the role of collaboration learning, We show that all agents gradually learn to behave as doves, which result in social efficiency.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) hawk-dove games / mutual learning / mimicry / local matching / co-evolution
Paper # AI2001-79
Date of Issue

Conference Information
Committee AI
Conference Date 2002/1/3(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Co-evolutions through Imitation Learning
Sub Title (in English)
Keyword(1) hawk-dove games
Keyword(2) mutual learning
Keyword(3) mimicry
Keyword(4) local matching
Keyword(5) co-evolution
1st Author's Name Yukikazu MURAKAMI
1st Author's Affiliation Dept. of Computer Science, National Defense Academy()
2nd Author's Name Hiroshi SATO
2nd Author's Affiliation Dept. of Computer Science, National Defense Academy
3rd Author's Name Akira NAMATAME
3rd Author's Affiliation Dept. of Computer Science, National Defense Academy
Date 2002/1/3
Paper # AI2001-79
Volume (vol) vol.101
Number (no) 536
Page pp.pp.-
#Pages 8
Date of Issue