De-Biasing The Lasso With Degrees-of-Freedom Adjustment

This paper studies schemes to de-bias the Lasso in a linear model where the goal is to construct confidence intervals for in a direction , where has iid rows. We show that previously analyzed propositions to de-bias the Lasso require a modification in order to enjoy efficiency in a full range of sparsity. This modification takes the form of a degrees-of-freedom adjustment that accounts for the dimension of the model selected by Lasso. Let be the true sparsity. If is known and the ideal score vector proportional to is used, the unadjusted de-biasing schemes proposed previously enjoy efficiency if . However, if , the unadjusted schemes cannot be efficient in certain : then it is necessary to modify existing procedures by a degrees-of-freedom adjustment. This modification grants asymptotic efficiency for any when and . If is unknown, efficiency is granted for general when \frac{s_0\log p}{n}+\min\Big\{\frac{s_\Omega\log p}{n},\frac{\|\Sigma^{-1}a_0\|_1\sqrt{\log p}}{\|\Sigma^{-1/2}a_0\|_2 \sqrt n}\Big\}+\frac{\min(s_\Omega,s_0)\log p}{\sqrt n}\to0 where , provided that the de-biased estimate is modified with the degrees-of-freedom adjustment. The dependence in and is optimal. Our estimated score vector provides a novel methodology to handle dense . Our analysis shows that the degrees-of-freedom adjustment is not needed when the initial bias in direction is small, which is granted under stringent conditions on . The main proof argument is an interpolation path similar to that typically used to derive Slepian's lemma. It yields a new error bound for the Lasso which is of independent interest.
View on arXiv