基于多Agent系统的自调节及协同工作的组合投资模型

Combination Investment Model of Auto-Adjustment and Cooperate with Work Based on Multi-Agent System

Agent技术特别是多Agent系统(MAS,Multi-Agent System)为解决人工智能等领域复杂问题提供了一个新途径,多Agent系统重点研究如何协调系统中的各个Agent的行为使其协同工作。针对多阶段组合投资问题,提出了一个基于多Agent系统的自调节及协同工作的组合投资策略模型。该模型系统中的各个Agent通过通讯共享知识,在求解问题的搜索空间中进行协同搜索,在更短的搜索步长内得到问题的解,极大地提高了系统性能。该模型具有不基于任何股票模型、时间复杂度低以及逼近最优投资策略速度较快等优点,实验证明具有一定的实际意义。

Agent technology,especially MAS（Multi-Agent System） provides a new method that can greatly resolve the complex problems in fields such as artificial intelligence,focus on multi-agent system,how to coordinate the various systems work together so that it acts Agent.In the paper,propose to the combination investment model of auto-adjustment and cooperation based on multi-agent system by multi-phases combination investment problem.In the model,every agent carry through cooperation search by the search space of the solution problems by communication sharing knowledge.The solution of the problems is gained by the shorter search step,and the system performance is maximum increased.The model has these excellence of the any stock model no relation,the low of time complexity and the speed fast of the approach optimal investment.It is signification from the experimentation show.