A Clustering Algorithm in Group Decision Making

The homogeneous requirement of the AHP has stifled its general application and thus hindered the further development of group decision making, which is becoming increasingly popular with the multi-business corporations. In this paper, we adopted the correlation degree of individual preference vectors as the measurement of individuals similarity based on AHP vector space VAHP proposed by Zahir in literature (Saaty 1989). The group in charge of decision making can thus be divided into multiple homogeneous clusters using this measurement, based on whose result the AHP can be applied. A heuristic algorithm is put forward at the end of the paper to facilitate the clustering of the group and the attainment of the group preference, both of which are vital to solve the problem of individual clustering in the group decision making process. Keyword: VAHP; Group decision making; Group Cluster, Group Coherence