A Hierarchical Clustering Approach to英文资料.pdfVIP

Proceedings of the Twenty-Third International Joint Conference on Artificial Intelligence C-Link: A Hierarchical Clustering Approach to Large-Scale Near-Optimal Coalition Formation∗ a a b a Alessandro Farinelli , Manuele Bicego , Sarvapali Ramchurn , Mauro Zucchelli aComputer Science Department, University of Verona, firstname.lastname@univr.it bSchool of Electronics and Computer Science, University of Southampton, sdr@ecs.soton.ac.uk Abstract To date, a number of approaches employed in this do- main aim to solve the CSG problem optimally, ranging Coalition formation is a fundamental approach to from Mixed-Integer programming to Branch-and Bound tech- multi-agent coordination. In this paper we address niques [Rahwan et al., 2009] through Dynamic Programming the specific problem of coalition structure gener- [ ] (DP) Yun Yeh, 1986; Rahwan and Jennings, 2008b . How- ation, and focus on providing good-enough solu- ever, none of these solutions are scalable given that the input tions using a novel heuristic approach that is based to the CSG problem is exponential in the number of agents on data clustering methods. In particular, we pro- (O(nn )). Hence these optimal approaches can handle rela- pose a hierarchical agglomerative