 Statistical Theory for High-Dimensional Models by Sara van de Geer

Primary text: arxiv.org/pdf/1409.8557

Cited passages: Sections 1.3–1.4, printed pp. 6–12: normalized Lasso and square-root Lasso objectives and prediction inequalities. Section 2.4, Theorems 2.4.1–2.4.2, printed pp. 18–19: linear-noise expansion, coefficient-error remainder and Gaussian normalization of de-sparsified Lasso.