We consider an independence feature screening technique for identifying explanatory variables that locally contribute to the response variable in high- dimensional regression analysis. Without requiring a specific parametric form of the underlying data model, our approach accommodates a wide spectrum of nonparametric and semiparametric model families. To detect the local con- tributions of explanatory variables, our approach constructs empirical likeli- hood locally in conjunction with marginal nonparametric regressions. Since our approach actually requires no estimation, it is advantageous in scenarios such as the single-index models where even specification and identification of a marginal model is an issue. By automatically incorporating the level of variation of the nonparametric regression and directly assessing the strength of data evidence supporting local contribution from each explanatory vari- able, our approach provides a unique perspective for solving feature screen- ing problems. Theoretical analysis shows that our approach can handle data dimensionality growing exponentially with the sample size. With extensive theoretical illustrations and numerical examples, we show that the local inde- pendence screening approach performs promisingly.
