Tuning parameter selection for penalized estimation via R2

The tuning parameter selection strategy for penalized estimation is crucial to identify a model that both interpretable and predictive. However, popular strategies (e.g., minimizing average squared prediction error via cross-validation) tend select models with more predictors than necessary. A simple yet powerful cross validation proposed which based on maximizing the correlation between observed predicted values, rather loss purposes of support recovery. can be applied all least-squares estimators and, under certain conditions, metric implicitly performs bias adjustment named α-modification. When Lasso estimator, α-modification closely related relaxed estimator. approach demonstrated functional variable problem optimal placement surface electromyogram sensors control robotic hand prosthesis.