As mining investment is a kind of venture capital,there are a lot of uncertainty and ambiguity in the process of evaluating on mining investment decision.With the data mining technology,the relationship of object classification and index membership is affirmed based on the entropy to mine knowledge information about object classification hidden in every index.The redundant data in index membership for object classification is eliminated by defining distinguishable weight and extracting valid values to compute object membership.The new algorithm of membership degree transformation consists of three calculation steps which could be summarized as "effective,comparison and composition",and denoted into M(1,2,3).The new algorithm is applied in fuzzy evaluation on mining investment decision,and the case study indicates that this method is characterized by practicability and high confidence for application.