Generalized Inverses, Ridge Regression, Biased Linear Estimation, and Nonlinear Estimation
Citations
7,165 citations
Cites methods from "Generalized Inverses, Ridge Regress..."
...Marquardt (1970) uses a VIF greater than 10 as a guideline for serious multi-collinearity....
[...]
3,697 citations
2,309 citations
2,290 citations
Cites methods from "Generalized Inverses, Ridge Regress..."
...In a forthcoming report we will also investigate statistical and numerical aspects on the PLS method, and show how the 8 theorems given by Marquardt (1970) for ridge and principal component regression translate into this situation....
[...]
2,147 citations
References
28,888 citations
"Generalized Inverses, Ridge Regress..." refers methods in this paper
...See [13, 17]....
[...]
...The antecedents of this work are: (1) my earlier work on algorithms for nonlinear estimation [17], where it was shown that a small positive quantity should be added to the diagonal of the linearized normal equations at each iteration, in order that the parameter correction vector can be confined to a region where the linearization yields meaningful approximations to the nonlinear function; (2) the work of A....
[...]
...The preceding theorems suggest that an algorithm (hereafter, the "alternate algorithm") for nonlinear estimation can be developed based on generalized inverses, and that this algorithm would have properties similar to those previously described [17] for the method (hereafter, the "original algorithm") based on ordinary inverses of (A + kI)....
[...]
8,300 citations
"Generalized Inverses, Ridge Regress..." refers background in this paper
..., by Rao ([22], [23], pp 183, 4), Golub [5, 6], and Goldman and Zelen [4]....
[...]
...Faced with the apparent singularity of A, it is tempting to invert A by means of a generalized inverse [7, 8, 22, 23]....
[...]
8,091 citations
5,182 citations
2,345 citations