利用非负约束主成份分析得出的一个解决校准统计问题的新方案

A NEW PROCEDURE OF MULTIVARIATE CALIBRATION BY USING NON-NEGATIVE CONSTRAINED PROGRAMMING AND MAIN COMPONENT ANALYSIS

本文对多维（Ｐ维）校准统计问题：Ｙｉ～Ｎｐ（αｘｉ＋β，Σ），利用主成份分析和非负约束规划的方法将原用反映指标向量Ｙ０估计相应之Ｘ０变为只用一个综合指标去估计Ｘ０，使多维校准问题降格成单维情况。（这里Ｙｉ和Ｙ０即反映指标向量，如α，β一样都是Ｐ维向量）

In this paper, we present a new procedure of multivariate calibration by using constrained programming and main component analysis, then reduced multivariate calibration ( P dimensions) problem to univariate case ( P =1).