解不确定型决策问题的投影寻踪方法

不确定型决策问题在系统工程理论与实践中具有重要意义 ,其益损值矩阵反映了该问题的复杂性和不确定性 ,同时也反映了决策者所面临的收益机会和损失机会的风险信息 ,目前常用方法利用益损值矩阵的信息都不充分 .指出求解该问题的实质 ,就是如何把益损值矩阵转换为压缩向量 ,该压缩向量的最大分量所对应的方案就是所求的最优方案 .为全面处理这种机会风险 ,提出了基于遗传算法的投影寻踪方法的新途径 .研究结果表明 ,根据益损值矩阵所反映的收益机会的风险与损失机会的风险的对比情况 ,投影寻踪方法既可积极地也可稳妥地选取最优方案 ,利用益损值矩阵的信息比常用方法充分 ,可根据决策者所面临的机会风险能进能退.

Projection Pursuit Method for Uncertain Decision-Making Problem

Uncertain decision-making problems take on very important meaning in theory and practice of systems engineering, their benefit and loss matrixes reflect high complexity and uncertainty of the problems, also reflect the risk information of benefit chance and loss chance faced by decision-maker. Now the information is not mined sufficiently by common methods for uncertain decision-making problems. The matter for resolving uncertain decision-making problem is how to transform benefit and loss matrix to compressed vector. The biggest weight of the vector corresponds to the best scheme. A new approach, named projection pursuit method based on genetic algorithm, is presented for utilizing the risk information completely. The result shows that decision-maker can actively or safely choose the best scheme based on the comparison between of benefit chance risk and loss chance risk contained in the benefit and loss matrix by using projection pursuit method.