标志空间的压缩是识别的一个重要问题.本文在K-L变换基础上,进一步考虑到均值的贡献,从而明显地提高了标志的识別能力.结合新疆托里金矿区,分析研究了综合标志的意义与作用.从最大隶属原则出发,给出了加权平均和模糊统计分类方法.最后,利用上述方法对该区作了金矿预测,预测结果与验证结果基本一致. The compression of sign space is an important problem in identification.This paper further considers the contribution of mean based on KL transformation, so as to obviously improve the identification ability of sign.Combined with the Tori gold mining area in Xinjiang, Meaning and function.According to the principle of maximum subordination, we give a weighted average and fuzzy statistical classification method.Finally, using the above method, the gold mine prediction is made in this area, and the prediction results are basically consistent with the verification results.