独立分量回归是一种将独立分量分析与回归分析相结合的方法,即先从自变量中提取出独立分量,然后利用独立分量进行回归建模。模拟实验表明,该方法既解决了各因子间的相关性问题,又能通过提取的独立分量对因变量进行更好的解释说明。
The independent component regression is a method that combines the independent component analysis with regression analysis, that extracts mutually independent components from the arguments first, and then builds the regression model with the independent components. The simulation experiment shows that this method can not only solve the problem of relativity of the impact factors, but also give a better explanation about the dependent variable.