Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§8.3.2, p. 70 — with AᵢᵀAᵢ = Q diag(λ) Qᵀ, ‖(AᵢᵀAᵢ + νI)⁻¹Aᵢᵀv‖₂ = ‖diag(λ + ν1)⁻¹QᵀAᵢᵀv‖₂

Proved
BoydADMM.ModelFit.group_lasso_eigen_norm

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

admmeigendecompositiongroup-lassop2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-bookp2o-v1

Let Ai∈Rm×niA_i\in\mathbb R^{m\times n_i}Ai​∈Rm×ni​ and v∈Rmv\in\mathbb R^mv∈Rm, and suppose

AiTAi=Q diag(μ) QTA_i^TA_i=Q\,\mathbf{diag}(\mu)\,Q^TAiT​Ai​=Qdiag(μ)QT

with Q∈Rni×niQ\in\mathbb R^{n_i\times n_i}Q∈Rni​×ni​ orthogonal and μ∈Rni\mu\in\mathbb R^{n_i}μ∈Rni​ (so μ\muμ is the vector of eigenvalues of AiTAiA_i^TA_iAiT​Ai​, the squares of the singular values of AiA_iAi​). Then for every ν>0\nu>0ν>0

∥(AiTAi+νI)−1AiTv∥2=∥diag(μ+ν1)−1QTAiTv∥2.\bigl\|(A_i^TA_i+\nu I)^{-1}A_i^Tv\bigr\|_2=\bigl\|\mathbf{diag}(\mu+\nu\mathbf 1)^{-1}Q^TA_i^Tv\bigr\|_2 .​(AiT​Ai​+νI)−1AiT​v​2​=​diag(μ+ν1)−1QTAiT​v​2​.

Once QTAiTvQ^TA_i^TvQTAiT​v is cached, the right-hand side costs O(ni)O(n_i)O(ni​) flops, which makes the one-parameter search over ν\nuν in the group-lasso update cheap.

Formalization Note The book calls the eigenvalue vector λ\lambdaλ, clashing with the regularization weight; here it is μ\muμ. The book assumes AiA_iAi​ tall (m≥nim\ge n_im≥ni​) for the cost count; the identity does not need it, and no such hypothesis is imposed. That μ\muμ consists of eigenvalues, and that they are nonnegative, follows from the decomposition, so it is not assumed separately.

Preamble
import Mathlib
import Definitions.Def_BoydADMM_ModelFit_Basic

open Matrix
Formal statement
namespace BoydADMM.ModelFit

/-- §8.3.2, p. 70: if `AᵢᵀAᵢ = Q diag(μ) Qᵀ` with `Q` orthogonal, then for `ν > 0`
`‖(AᵢᵀAᵢ + νI)⁻¹Aᵢᵀv‖₂ = ‖diag(μ + ν1)⁻¹ Qᵀ Aᵢᵀ v‖₂`. (The book's eigenvalue vector `λ` is `μ`.) -/
theorem group_lasso_eigen_norm {m n : ℕ} (A : Matrix (Fin m) (Fin n) ℝ)
    (v : EuclideanSpace ℝ (Fin m)) (Q : Matrix (Fin n) (Fin n) ℝ)
    (hQ : Q ∈ Matrix.orthogonalGroup (Fin n) ℝ) (μ : Fin n → ℝ)
    (hdecomp : Aᵀ * A = Q * Matrix.diagonal μ * Qᵀ) (ν : ℝ) (hν : 0 < ν) :
    ‖ridgeSol A ν v‖ =
      ‖Matrix.toEuclideanLin ((Matrix.diagonal (fun j => μ j + ν))⁻¹ * Qᵀ * Aᵀ) v‖ := by sorry

end BoydADMM.ModelFit
Source
Boyd, Parikh, Chu, Peleato, Eckstein, Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers, Found. Trends Mach. Learn. 3(1) (2011), p. 70, §8.3.2
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me