A Rational Interpretation of the Ridge Trace

From MaRDI portal
Publication:3217463





The ridge regression estimator is expressed as a weighted average of least squares estimators of all possible subset regressions and is exploited to understand the behaviour of the ridge trace asymptotically as well as near \(k=0.\) Through Bayes formulation, the weights are shown to incorporate sample information and prior information. This approach might help in understanding the sensitivity of regression estimates to the choice of variables in the model through the ridge trace. Insights achievable are illustrated through a real problem with ten explanatory variables.











This page was built for publication: A Rational Interpretation of the Ridge Trace

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3217463)