一种区间数的因子分析技术及其在证券市场中的应用

传统的因子分析技术能够有效地对高维变量空间进行降维处理,但它对于样本空间却缺乏行之有效的降维效果.为了解决这一问题,一种针对大量样本数据、新的因子分析技术———区间数因子分析技术(intervaldatafactoranalysis,IFA)被提出并得到了迅速的发展。IFA方法对传统的数据概念做了本质性的扩张,运用'数据打包'的理念,对海量原始数据在不破坏其原有内在逻辑关系的前提下,可以进行变量和样本点维度的双重降维。本文详细阐述了区间数因子分析技术的原理,并以中国股票市场为案例研究背景,结果表明IFA分析技术对大规模多维数据系统做综合简化是十分有效的。

An Interval Data Factor Analysis Method And Its Application

A new kind of advanced complicated data analysis technology-Interval Data Factor Analysis (IFA)-has been brought forward and used in this paper. Compared with traditional factor analysis technology, the IFA technology can actualize double decreasing dimension, both in proxy dimension and in observation dimension. This paper has introduced the principal of IFA technology detailedly, and regarded style indices of Chinese international trust & investment company as the research object. The research result indicates that the style characteristic is disinct in Chinese stock market, which, as well as the factor of scale, has influenced Chinese stock market dramatically. In addition, the paper states that different style indices have remarkable difference in behavior characteristics, such as the Risk-Return quality. The computational analysis has showed that the research result is very consistent with the realistic characteristics of Chinese stock market, which proves that it is very effectual to simplify the multidimensional dynamic data system, with the help of IFA method.