Superior information is insufficient to win in games between finite automata

被引:0
作者
Chernorutskii, V
Izmailov, R
Pokrovskii, A
机构
[1] NEC USA INC,C&C RES LABS,PRINCETON,NJ 08540
[2] UNIV QUEENSLAND,DEPT MATH,BRISBANE,QLD 4072,AUSTRALIA
关键词
repeated games; finite automata; information; games on graphs;
D O I
10.1137/S0363012992239892
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
A game between two computers is considered: the first computer generates a binary sequence while the second one tries to predict the next element of this sequence using the previous elements. Both computers operate with the same pool of strategies, which is the set of all boolean functions of N arguments. Notwithstanding the asymmetry of the game, it turns out that the value of the game is zero. An algorithm for choosing an optimal superstrategy for the first computer is proposed, and several generalizations of the game are considered.
引用
收藏
页码:542 / 553
页数:12
相关论文
共 11 条
[1]   THE STRUCTURE OF NASH EQUILIBRIUM IN REPEATED GAMES WITH FINITE AUTOMATA [J].
ABREU, D ;
RUBINSTEIN, A .
ECONOMETRICA, 1988, 56 (06) :1259-1281
[2]   GAMES WITH REPEATED DECISIONS [J].
ALPERN, S .
SIAM JOURNAL ON CONTROL AND OPTIMIZATION, 1988, 26 (02) :468-477
[3]  
[Anonymous], 1982, GAME THEORY
[4]  
BERGE C, 1991, N HOLLAND MATH LIB 1, V6
[5]   FINITE RATIONALITY AND INTERPERSONAL COMPLEXITY IN REPEATED GAMES [J].
KALAI, E ;
STANFORD, W .
ECONOMETRICA, 1988, 56 (02) :397-410
[6]  
OPOYTSEV VI, 1993, AUTOMAT REM CONTR, V7, P201
[7]  
Pokrovskii A. V., 1989, Soviet Physics - Doklady, V34, P594
[8]  
TSETLIN ML, 1963, SOV MATH DOKL, V149, P284
[9]  
TSETLIN ML, 1963, SOV MATH DOKL, V149, P52
[10]  
VLADIMIROV AA, 1982, T VNIISI GAME DYNAMI, V4, P84