中文摘要：

在多指标不确定决策中，如果选用隶属函数描述各维指标值关于状态的不确定性，必然涉及坐标隶属度到点隶属度的转换问题．现有的隶属度转换方法受到冗余数据的干扰，影响了决策结果的可信性．为此，通过挖掘隐藏在坐标隶属度中关于点分类的知识信息，厘清点分类与坐标隶属度间的关系；定义指标区分权清除坐标隶属度中对点分类不起作用的冗余值，构建从坐标隶属度到点隶属度转换的数学方法，为多指标不确定决策提供方法依据．

英文摘要：

In any multi-indexes uncertainty decisions, if choosing membership function to de-scribe the uncertainty of every space-index value about some status, it must meet membershipdegree conversion, which must solve the conversion from membership degree of coordinatesto membership degree of points. But the present nlethods effect the credibility of result ofdecision because of the redundant data. So, by mining knowledge information about points＇classification hidden in membership degree of every coordinates, affirming the relation between points＇ classification and membership degree of coordinates; eliminating the redundantdata in coordinates＇ menlbership for points＇ classification by defining distinguishable weight;constructing the mathematic nlethod of conversion from membership degree of coordinates tomembership degree of points without the interference of redundant data, providing methodbasis for solving key computation in multi-indexes uncertainty decisions.