Robust model-free feature screening via quantile correlation

In this paper, we propose a new sure independence screening procedure based on quantile correlation (QC-SIS). The method not only is robust against outliers, but also can discover the nonlinear relationship between independent variables and dependent variable. We establish the sure screening property under certain technical conditions. Simulation studies are conducted to assess the performances of QC-SIS, sure independent screening (SIS), sure independent ranking and screening (SIRS), robust rank correlation screening (RRCS) and distance correlation-sure independent screening (DC-SIS). Results have shown the effectiveness and the flexibility of the proposed method. We also illustrate the QC-SIS through an empirical example.