Skip to content

Adaptive Covariate Selection Under Dependence

Home / Open Questions / Adaptive Covariate Selection Under Dependence

Background: The selection of an optimal regressor index in high-dimensional marginal regression settings often necessitates adaptive post-model-selection inference.

Question / Future Work: The development of an endogenously selected optimal covariate index remains to be explored under weak dependence. This is particularly relevant for constructing tests that avoid the limitations of current pre-estimation screening methods while maintaining robust size and power properties.

Why It Matters: Developing such an index would bridge the gap between simple marginal screening and complex post-model-selection inference, potentially enhancing the detection of informative predictors in dependent data.

Metadata & Links