多元质量特性预报：MULTIVARIATE回归分析的应用

对现象之间客观存在的因果关系建立回归分析模型,这是实际中较为普遍的做法.在这篇文章中,我们根据MULTIVARIATE回归分析的基本原理,利用从生产现场采集的观测数据,对产品两个质量特性及其五个关键影响因素之间的关系建立了多重多元回归分析方程,为说明MULTIVARIATE回归应用的可行性,我们还结合实例给出了因变量向量估计的两种形式,以及无条件预报的置信区间。

Multivariate Regression Prediction Analysis of Multiple Quality Characteristics

The evolution of regression analysis,that is,from simple regression analysis to multiple regression analysis,and then multivariate regression analysis,is the inevitable outcome of theoretical study and practical needs.In this paper,with the employment of observed data collected from actual production process,we set up a multivariate regression analysis model to examine the relationship among two quality characteristics and five key influencing factors of the product.Furthermore,we make the unconditional prediction analysis of product quality characteristics based on it.