167
v1v2 (latest)

Exact variable selection in sparse nonparametric models

Electronic Journal of Statistics (EJS), 2023
Abstract

We study the problem of adaptive variable selection in a Gaussian white noise model of intensity ε\varepsilon under certain sparsity and regularity conditions on an unknown regression function ff. The dd-variate regression function ff is assumed to be a sum of functions each depending on a smaller number kk of variables (1kd1 \leq k \leq d). These functions are unknown to us and only few of them are nonzero. We assume that d=dεd=d_\varepsilon \to \infty as ε0\varepsilon \to 0 and consider the cases when kk is fixed and when k=kεk=k_\varepsilon \to \infty, k=o(d)k=o(d) as ε0\varepsilon \to 0. In this work, we introduce an adaptive selection procedure that, under some model assumptions, identifies exactly all nonzero kk-variate components of ff. In addition, we establish conditions under which exact identification of the nonzero components is impossible. These conditions ensure that the proposed selection procedure is the best possible in the asymptotically minimax sense with respect to the Hamming risk.

View on arXiv
Comments on this paper