Analysis of Variance, ANOVA
摘要
A function \(f(\textbf{x})\) , assumed depending on D random variables, \(x_1, \dots , x_D\) , can be decomposed as [1] that, with the property of with \(\mathbb E_i\) the expectation with respect to any coordinate i in the set \((i_1, \dots , i_d)\) , \(1 \le d \le D\) , results in the orthogonality of functions involved in the previous decomposition.