Lasso Tobit Principal Component Regression With An Application

One of the most crucial subjects in the analysis of statistical models is the identification of important variables. Therefore, the search for best variable selection methods is a good in obtaining best estimators. The Lasso method is considered the most effective approach for variable selection and parameter estimation in building statistical models with high explanatory power in representing the studied phenomenon. Therefore, using the Lasso method to estimate the parameters of a regression model that contains a dependent variable with data that is censored at zero can be achieved through the use of Lasso tobit principal component regression, it has attractive properties in estimating the parameters of this model. The our proposed method is illustrated via simulation scenario and a new real data .