中心化順序統計量與其應用

Abstract

本論文所探討的是關於多變量中心化順序統計量與其應用。首先，我們定 義多變量順序統計量再將其應用於多變品質管制，及建立多變量回歸截斷 平均數。最後，我們將多變量回歸截斷平均數與幾個估計方法做比較，由 結果可看出我們所建立的多變量回歸截斷平均數是個有效率的估計方法。 This thesis is concerned with the multivariate centered order statistics and its applications. First, we give the definition of multivariate centered order statistics. Then we apply it to multivariate quality control and set up a multivariate regression trimmed mean. Finally, the multivariate regression trimmed mean is compared with several estimation methods. From the comparision, we see that the trimmed mean which we have given is an efficient estimator for the multivariate regression parameter.