上海交通大学学报(自然版) ›› 2013, Vol. 47 ›› Issue (04): 644-649.

• 自动化技术、计算机技术 • 上一篇    下一篇

基于分类器融合的自动化协商决策模型

 彭艳斌1, 郑志军1, 于成波2, 李吉明3   

  1. (1.浙江科技学院 信息与电子工程学院, 杭州 310023; 2.浙江大学 计算机科学与技术学院,杭州 310027; 3.浙江警察学院 刑事科学技术系, 杭州 310053)
  • 收稿日期:2012-05-02 出版日期:2013-04-28 发布日期:2013-04-28
  • 基金资助:

    国家自然科学基金(61175058), 浙江省自然科学基金(Y1100036,LY12F02018), 浙江省教育厅科研计划基金(Y201016929,Y201222997),浙江省自然科学基金(Z12F020019)资助项目

Automated Negotiation Decision Model Based on Classifier Fusion

 PENG  Yan-Bin-1, ZHENG  Zhi-Jun-1, YU  Cheng-Bo-2, LI  Ji-Ming-3   

  1. (1. School of Information and Electronic Engineering, Zhejiang University of Science and Technology, Hangzhou 310023, China; 2. School of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China; 3. Department of Forensic Science, Zhejiang Police College, Hangzhou 310053, China)
  • Received:2012-05-02 Online:2013-04-28 Published:2013-04-28

摘要: 为了解决电子商务环境中由于信息的保密性使协商参与者无法获得对手协商偏好从而影响协商性能的问题,提出一种基于分类器融合的自动化协商决策模型.该模型融合支持向量机和贝叶斯分类器,通过结合2种分类器的优点,提高对协商偏好的分类学习效果.在准确估计对手协商偏好的基础上,采用粒子群优化算法搜寻最优协商反建议.实验数据分析表明,新方法的效果优于单一分类器,并且在有噪声的小规模训练样本集下,仍然保持较高的协商总效用.    

关键词: 自动化协商, 支持向量机, 贝叶斯

Abstract: Due to the confidentiality of information in e-commerce environment, negotiation participants can not get opponent’s negotiation preferences, thereby affecting the negotiation performance. To solve this, an automated negotiation decision model based on classifier fusion was proposed. The model incorporates support vector machine and Bayesian classifier by combining the advantages of both, improving the effect of classification learning of negotiation preferences. Based on accurate estimation of opponent’s negotiation preference, a particle swarm optimization algorithm was used to search the optimal counter proposal. The experimental data show that the new method is better than the single classifier, and can maintain a high total negotiation utility in the noisy small scale training set.  

Key words: automated negotiation, support vector machine, Bayesian

中图分类号: