Generalized Ridge Regression: Applications to Nonorthogonal Linear Regression Models
By: Román Salmerón Gómez, Catalina García García, Guillermo Hortal Reina
Potential Business Impact:
Fixes math problems when numbers are too similar.
This paper analyzes the possibilities of using the generalized ridge regression to mitigate multicollinearity in a multiple linear regression model. For this purpose, we obtain the expressions for the estimated variance, the coefficient of variation, the coefficient of correlation, the variance inflation factor and the condition number. The results obtained are illustrated with two numerical examples.
Similar Papers
Finite-Sample Properties of Generalized Ridge Estimators in Nonlinear Models
Methodology
Improves computer predictions by balancing accuracy and simplicity.
Stepwise regression revisited
Methodology
Fixes math models that have too many numbers.
Generalisation and benign over-fitting for linear regression onto random functional covariates
Machine Learning (Stat)
Helps computers learn from messy, connected data.