基于独立成份分析的证券市场分析  

A Method Based on Independent Component Analysis Used for Stock Market Analysis

作　　者：马裕[1] 段哲民[1] 彭斌[1] 霍建[1] 

机构地区：[1]西北工业大学电子信息学院,陕西西安710072

出　　处：《计算机仿真》2009年第10期323-326,共4页Computer Simulation

摘　　要：为研究独立成分分析用于证券市场分析,首先介绍了独立成分分析的理论基础,然后介绍了一种独立成分分析的具体算法Fastica算法。为分析出对股票成交量影响的因素,进行了几只重点股票交易额的ICA分离实验,并对实验结果进行了分析研究,从而找到了影响股票交易量的几个重要因素,揭示出了股票交易额变化的几个重要原因。结果说明ICA是揭示隐藏在想象背后的原因的有效手段,非常适合于对证券市场的分析。To study the independent component analysis for stock market analysis, the paper first introduced the theory of independent component analysis , then introduced a specific algorithm--Fastica algorithm. In order to analyze the factors influencing the stock turnover, the ICA separation experiments were performed for several stocks＇ turnover, and experimental results are analyzed tto find some important factors which affect the volume of stock transactions. The reaslt demonstrated that ICA is an effective means to reveal the hidden reasons, and is very suitable for the analysis of the stock market.

关 键 词：独立成份分析 交易额 证券市场 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]