Coalition Calculation in a Dynamic Agent Environment
Ted Scully - National University of Ireland, Galway Michael G. Madden - National University of Ireland, Galway Gerard Lyons - National University of Ireland, Galway
We consider a dynamic market-place of self-interested agents with differingcapabilities. A task to be completed is proposed to the agent population. Anagent attempts to form a coalition of agents to perform the task. Beforeproposing a coalition, the agent must determine the optimal set of agents withwhom to enter into a coalition for this task; we refer to this activity ascoalition calculation. To determine the optimal coalition, the agent must havea means of calculating the value of any given coalition. Multiple metrics(cost, time, quality etc.) determine the true value of a coalition. However,because of conflicting metrics, differing metric importance and the tendencyof metric importance to vary over time, it is difficult to obtain a truevaluation of a given coalition. Previous work has not addressed these issues.We present a solution based on the adaptation of a multi-objectiveoptimization evolutionary algorithm. In order to obtain a true valuation ofany coalition, we use the concept of Pareto dominance coupled with a distanceweighting algorithm. We determine the Pareto optimal set of coalitions andthen use an instance-based learning algorithm to select the optimal coalition.We show through empirical evaluation that the proposed technique is capable ofeliciting metric importance and adapting to metric variation over time.