Penalized LAD-SCAD Estimator Based on Robust Wrapped Correlation Screening Method for High Dimensional Models

The widely used least absolute deviation (LAD) estimator with the smoothly clipped (SCAD) penalty function (abbreviated as LAD-SCAD) is known to produce corrupt estimates in presence of outlying observations. problem becomes more complicated when number predictors diverges. To overcome these problems, LAD-SCAD based on sure independence screening (SIS) technique put forward. SIS method uses rank correlation (RCS) algorithm pre-screening step and traditional Pathwise coordinate descent for computing sequence regularization parameters post onward model selection. It now evident that less robust against outliers. Motivated by inadequacies, we propose improvise using wrapped (WCS) replacing correlation. proposed denoted WCS+LAD-SCAD will be employed variable simulation study real-life data examples show procedure produces efficient results compared existing methods.