基于矢量空间的群体聚类方法研究

通过对群体成员的思维模式进行分析,基于矢量空间提出了群体中任意两个成员偏好矢量相关性度量的一般范数以及基于该矢量相关度的启发式算法,通过执行该使算法使群体中的成员形成若干个不同的聚集。在此基础上,定义并计算群体中各个聚集的偏好和整个群体的偏好,定义并计算群体中各个聚集的一致性指标和整个群体的一致性指标,为群体决策提供依据。最后通过一个实例对算法和一致性指标等进行验证。

Research on the group clustering method based on vector space

By analyzing the thinking pattern of group members, the general norm of relativity measurement of preference vectors of any two members in the group and the heuristic algorithm based on this vector relativity degree are put forward based on vector space. With the algorithm, several different clusters are formed in the group. On this basis, the preferences of every cluster of the group and the preference of the entire group, as well as the coherence indexes of every cluster of the group and the coherence index of the entire group are defined and calculated, which provides the basis for group decision. Finally, the algorithm and coherence indexes proposed are validated by an example.