利用K-L变换提高模式识别的精度  

Pattern recognition improvement using K-L transform

作　　者：吴大奎[1] 

机构地区：[1]成都理工学院,610059

出　　处：《石油地球物理勘探》1995年第2期272-276,共5页Oil Geophysical Prospecting

摘　　要：本文采用直接方式和K－L变换方式分别对地震反射信息确定的样本进行了分类。其结果表明：K－L变换前，分类能力不强，而K－L变换后，分辨能力有很大提高，可以更准确地提取到所需样本。此外，K－L变换前，运算速度慢、收敛慢（此处的收敛是指本文以迭代自组织方法聚类时达到需要的类别数）；而K－L变换后，运算速度快、收敛也快，且样本数、变量数越多，其优越性越明显。The samples of seismic reflection informations are clustered by using direct way and K-L trans form respectively. It can be seen that the clustering is poor before K-L transform, but irnproves greatly after K-L transform to get better samples we need. What is more, K-L transform makes computation and convergence faster than before (The convergence means that we come up to an ideal category number in the case of iterative self-organization clustering)- The more the samples and variables, the more superior the result.

关 键 词：K-L变换 模式识别 地震勘探 数据处理 

分 类 号：P631.443[天文地球—地质矿产勘探]