Abstract
In order to study bounded rationality behavior in experimental economics, a simulation platform based on finite automaton for multi-agent was proposed. In the frame, the behavior of organization and individual was expressed by finite automaton that was constructed by object-oriented visual interface technology. The finite automaton was translated into thread with translating language in artificial intelligence, and then the threads were processed through concurrent and serial schedule way to realize sequential game, mixed game and evolutionary game. In the game process, the new finite automaton was generated by genetic algorithm. The evolutionary process in three-person iterated prisoner's dilemma game was simulated in the end.
Original language | English |
---|---|
Pages (from-to) | 6658-6663 |
Number of pages | 6 |
Journal | Xitong Fangzhen Xuebao / Journal of System Simulation |
Volume | 20 |
Issue number | 24 |
Publication status | Published - 20 Dec 2008 |
Externally published | Yes |
Keywords
- Experimental economics
- Finite automaton
- Genetic algorithm
- Java language
- Repeated game