基于agent的计算金融中agent的适应性模型

The Model of Agent Adaptation for Agent-based Computational Finance

在有关基于agent计算金融的研究中,主要强调agent的学习进化机制,却忽略了对agent使用相应学习机制的内在原因—agent不同适应属性的研究。在现实的金融市场中,各个agent在可用信息的解释、理解,本身的认知结构,对风险的态度,时间范围的认识,及决策规则等方面都存在相当大的差异。

Many current studies on agent -based computational finance focuses on learning and evolution. However, they ignore the intrinsic cause of agent learning - agent adaptation. The paper proposes a hiherarchy framework of agent adaptation based on agent behavior model. It discuss various mechanisms for adaptation occurring agent - based computational finance and explains the significance of exploring agent adaptation in multi agent simulation of stock market.