Olsavs: A New Algorithm For Model Selection

The shrinkage methods such as Lasso and Relaxed introduce some bias in order to reduce the variance of regression coefficients multiple linear models. One way after would be apply ordinary least squares subset predictors selected by method used. This work extensively investigated this idea developed a new variable selection algorithm. authors named technique OLSAVS (Ordinary Least Squares After Variable Selection). algorithm was implemented R. Simulations were used illustrate that is able produce better predictions with less for various error distributions. compared few widely terms their achieved test root mean square bias.