Coalition structure generation in characteristic function games

Acknowledgements I would like to first thank my advisor Dr. Julie Adams for her guidance and support of my research as well as the effort and hours she expended reading and correcting drafts of my work. I would also like to thank all of my committee members: Dr. Gautam Biswas, in general also like to thank all of my friends and family. In particular, I would like to thank my fiancee, Robyn, for being supportive throughout the past four years. IV.2 Approximation ratios and corresponding running times for the determinis-tic and randomized approximation algorithms for arbitrary coalitional games. 57 IV.3 Approximation ratios and corresponding run time of the presented mono-V.1 The average utilities generated by the randomized and deterministic algorithms for five different approximation ratios on three problem distributions V.2 The average utilities generated by the IP algorithm and the deterministic 2 3-approximation algorithm on the Uniform, Normal, Modified Uniform, and NDCS problem distributions consisting of 24 agents. The utility for the IP algorithm is averaged over the utility of the current best known coalition structure at the moment when the IP algorithm guarantees a 2 3-approximate V.3 The average utilities generated by Sandholm's algorithm, Dang and Jen-ning's algorithm and the deterministic approximation algorithm for 1 3-and 1 2-approximation ratios on the problem distributions for 17 agents. The utility for Sandholm's and Dang and Jenning's algorithms is averaged over the utility of the current best known coalition structure at the moment when each algorithm guarantees the stated approximation