线性模型参数的聚类主体成分估计

Clustering Principal Component Estimationof Linear Model Parameters

结合聚类分析方法与主成分分析方法，得到了线性模型参数的一种新估计聚类主成分估计，并把它与最小二乘估计（ＬＳＥ）和主成分估计（ＰＣＥ）比较，获得了一些优良的性质．

Combining clustering method with principal component estimate,a new estimationclustering principal component estimation of linear model parameters is obtained.Meantime,comparing it with other two estimates(LSE and PCE),its merits are pointed out.