Incorporating Bayesian learning in agent-based simulation of stakeholders’ negotiation

摘要: abstract This paper describes the incorporation of a Bayesian learning algorithm into an agent-based modeldesigned to simulate stakeholders’ negotiation when evaluating scenarios land development. Theobjective is facilitate reaching agreement at earlier stage in by providing theopportunity proposer agent learn his opponents’ preferences. The modeling approach testedin Elbow River watershed, southern Alberta, Canada, that under considerable pressure for landdevelopment due proximity fast growing city Calgary. Five agents are included themodel respectively referred as Developer agent, Planner Citizen Agriculture-Concerned and WaterConcerned agent. Two types development evaluated;in first case, only geographical location considered while second internal land-use composition also varied. equipped with capabilityattempts approximate its fuzzy evaluation functions based on responses he receivesfrom them each round negotiation. results indicate using this approach, agreementcan be reached fewer number rounds than case where selectsthe subsequent offers merely own utility. model indicates how satisfaction ofeach evolves during information very useful decision makers who wishto consider perspectives dealing multiple objectives spatial context. 2014 Elsevier Ltd. All rights reserved.