Multi-Objective Oriented Categorization Based on the Coalitional Game Theory
Multi-Objective Oriented Categorization Based on the Coalitional Game Theory
复制标题
基于联盟博弈论的多目标定向分类
DOI：
10.1142/s0218213016500111
复制
发表时间：
2016
影响因子：
1.1
通讯作者：
Li Jin
中科院分区：
文献类型：
--
作者：
Liu Weiyi;Yue Kun;Fu Xiaodong;Yin Zidu;Li Jin
发现不同的组或称为类，对于模式识别、数据预处理、关联分析、查询优化等都是有用的。为了使每个对象尽可能地满足，组是由参与对象之间的关联或行为生成的，而不是由参与对象本身拥有的属性生成的。本文以博弈论为基础，主要考虑给定对象之间的相互关联，研究面向多目标的分类问题。基于联盟博弈中Shapley值的思想，首次提出了优先级组的概念，并给出了计算组内成员满意度的有效算法。基于战略博弈的思想和纳什均衡的思想，给出了求解近似均衡的算法，以解决博弈双方策略之间的冲突，从而实现最终的多目标导向群体。初步实验和性能研究验证了该方法的有效性和有效性。
Discovering different groups, or called classes, is useful for pattern recognition, data preprocessing, association analysis, query optimization, etc. To make every object satisfied as much as possible, the groups are generated by the associations or behaviors among participating objects other than the attributes owned by themselves. By mainly considering the mutual associations among the given objects and based on the game theory, in this paper we study the multi-objective oriented categorization. Based on the idea of Shapley value in the coalitional game, we first propose the concept of priority groups and give the efficient algorithm for computing the satisfaction degree of players in a group. Based on the idea of strategic games and Nash equilibrium, we then give the algorithm for computing an approximate equilibrium to solve the conflicts between the strategies of players, and consequently achieve the ultimate multi-objective oriented groups. Preliminary experiments and performance studies verify the efficiency and effectiveness of our methods.