COLLABORATIVE ORGANIZATIONAL LEARNING ALONG A DIRECTION

摘要

We introduce the effectiveness of agents as vectors relative to the desired direction for improvement and define a coalition as a collaboration of agents that can improve individual values based on a utility function. We develop algorithms for finding sequentially formed winning coalitions without and with repeated hiring by which an agent joins the coalition if it can improve the values for all existing agents. An algorithm for finding the optimal sequentially formed efficient core is also proposed.