影响图及遗传算法的私人信息推测

市场竞争中,准确推测竞争对手的私人信息对于竞争者掌握竞争的主动权具有重要意义。以影响图作为竞争对手的决策模型,以对手的目标是使整体决策的总期望效用最大,而不仅仅是使每一步决策获得最大效用为准则,根据所观测到的竞争对手的决策行为,利用遗传算法推测对手的自然信息、决策信息或效用信息。实验结果表明,所提方法是正确有效的。

Influence diagrams and genetic algorithm-based rival's private information inferring.

It is important for having the initiative in competition to infer the rival’s private information accurately in the competition.The rival’s decision-making model is described by Influence Diagrams（IDs）.According to the criterion that rival is apt to gain the maximum expected utility in whole multi-steps decision rather than gain the maximum expected utility just in each of the decision steps，the rival’s private information about cost，bankroll，tactic，and utility can be inferred through the observation on rival’s decision behaviors and using the Genetic Algorithm（GA）.The experimental results show that proposed method is both accurate and reasonable.