K—L变换地震属性优化方法

本文研究K—L（Karhuen—Loeve）变换方法（主成分分析），井应用常规K—L变换对原始地震属性进行处理。从全体地震属性集中，挑选最好的地震属性子集以降低多解性。去除相关属性。重构新的地震属性。

This paper introduces the K-L （Karhuen-Loeve） transform method （main composition analyze）, and how to use the reguLar K-L transform to process the primitive earthquake attribute in order to select the subset of best earthquake attribute from whole earthquake attribute to reduce the multi-solutions, remove the correlation attribute, reconstruct the new attribute.