摘要
本文针对传统多因素模型的两个重要步骤:模型形式设定和变量选择问题,结合我国股市的特点,利用前馈多层神经网络和层次贡献分析方法相结合,建立了影响我国股市的宏观多因素模型,并进行了实证研究,证明了该方法的有效性。
There exist two important processes in the traditional Multi-factor model. One is to decide the model’s function form, the other is to decide which variables to be put into the model. According to the characteristics of the Chinese emerging stock market, we combine the technology of neural network and hierarchical contribution analysis to make a new multi-factor model that is more suitable for the Chinese stock market. The result of its application demonstrates its efficiency.
出处
《武汉船舶职业技术学院学报》
2005年第2期63-67,共5页
Journal of Wuhan Institute of Shipbuilding Technology
关键词
素模型
前馈神经网络
股市
宏观
前馈多层神经网络
选择问题
实证研究
变量
portfolio investment
multi-factor model
yield
feed-forward multi-layer neural network
hierarchical contribution analysis.