文中提出了一种预测协商中Agent行为的学习机制，该机制的基础是仅使用协商交往中对方的历史响应进行非线性回归分析。自动协商中对方Agent的行为由其决策函数表示的策略决定。先通过一系列的模拟得到对方在采用各种策略和参数配置的响应，然后总结提取了估计对方策略的启发性知识，最后把此知识应用到实验性的在线协商中进行测试。结果表明使用这些知识能够取得比现有决策函数策略更好的结果。该学习机制可以在线使用，也不需要有关于对方的过去知识，在双方不了解或很少了解的开放式系统中尤为有效。%A learning mechanism to predict a negotiation Agent's behaviour is presented,the basis of the mechanism is only to apply the opponent's previous offers for nonlinear regression analysis. The behaviour of negotiation Agents is determined by their tactics in the form of decision functions. Heuristics based on estimates of an Agent's tactics are drawn from a series of experiments with varying tactics and combinations of parameters. The obtained heuristics is then applied to a series of simulated online negotiation sessions. The results of these simulated sessions show that this approach can be used to obtain better deals than existing decision function tactics. The learning mechanism can be used online,without any prior knowledge about the other Agents,therefore is very useful in open systems where Agents have little or no information about each other.
展开▼
