《多元统计分析》课程简介

研究客观事物中多个变量（或多个因素）之间相互依赖的统计规律性。它的重要基础之一是多元正态分析。又称多元分析 。 如果每个个体有多个观测数据，或者从数学上说， 如果个体的观测数据能表为 P维欧几里得空间的点，那么这样的数据叫做多元数据，而分析多元数据的统计方法就叫做多元统计分析 。 它是数理统计学中的一个重要的分支学科。20世纪30年代，R.A.费希尔，H.霍特林，许宝碌以及S.N.罗伊等人作出了一系列奠基性的工作，使多元统计分析在理论上得到迅速发展。50年代中期，随着电子计算机的发展和普及 ，多元统计分析在地质 、气象、生物、医学、图像处理、经济分析等许多领域得到了广泛的应用 ，同时也促进了理论的发展。各种统计软件包如SAS，SPSS等，使实际工作者利用多元统计分析方法解决实际问题更简单方便。重要的多元统计分析方法有：多重回归分析（简称回归分析）、判别分析、聚类分析、主成分分析、对应分析、因子分析、典型相关分析、多元方差分析等。

学时：48学时

Course Introduction to multivariate statistical analysis

Research objective things in multiple variables (or more) of the interdependence between statistical regularity. It is one of the important basis of multivariate normal analysis. Also known as multivariate analysis. If every individual have more than one observation data, or mathematically speaking, if the individual observations can table for P dimensional Euclidean space, then the data is called a multivariate data, and analysis of the multivariate statistical method is known as multivariate statistical analysis of data. It is an important branch in mathematical statistics. In the 1930 s, R.A. fisher, h. hotelling, Xu Bao the importance and the S.N. Roy and others made a series of foundational work, make the multivariate statistical analysis in theory get rapid development. 50 s, with the development and popularization of electronic computer, multivariate statistical analysis in geology, meteorology, biology, medicine, image processing, economic analysis and many other fields has been widely used, but also promote the development of the theory. Various statistical software packages such as SAS, SPSS, etc., make the practitioners using multivariate statistical analysis methods to solve practical problems is more easy and convenient. Important of multivariate statistical analysis methods are: the multiple regression analysis (regression analysis), discriminant analysis, cluster analysis and principal component analysis, correspondence analysis, factor analysis, canonical correlation analysis and multivariate analysis of variance, etc.

Time:48 class hours