Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 3 — the violation set V of Algorithm 2 is a correlation threshold test

Proved
L0BnB.DualGap.proposition_3

by mikedeng1 · Oct 8, 2026 · Mathlib 0df444a (Lean v4.33.1)

active-setoptimality-conditionsp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1sparse-regression

Let X∈Rn×pX\in\mathbb R^{n\times p}X∈Rn×p have columns of unit ℓ2\ell_2ℓ2​ norm, y∈Rny\in\mathbb R^ny∈Rn, λ0,λ2,M>0\lambda_0,\lambda_2,M>0λ0​,λ2​,M>0, and let β^∈Rp\hat\beta\in\mathbb R^pβ^​∈Rp with residual r^=y−Xβ^\hat r=y-X\hat\betar^=y−Xβ^​. Then the set VVV of Step 2 of Algorithm 2 (the coordinates i∉Supp⁡(β^)i\notin\operatorname{Supp}(\hat\beta)i∈/Supp(β^​) at which 000 is not a minimizer of βi↦F(β^1,…,βi,…,β^p)\beta_i\mapsto F(\hat\beta_1,\dots,\beta_i,\dots,\hat\beta_p)βi​↦F(β^​1​,…,βi​,…,β^​p​) over ∣βi∣≤M|\beta_i|\le M∣βi​∣≤M) can be written as

V={ i∈Supp⁡(β^)c  :  ∣⟨r^,Xi⟩∣>c(λ0,λ2,M) },(17)V=\{\,i\in\operatorname{Supp}(\hat\beta)^c \;:\; |\langle\hat r,X_i\rangle|>c(\lambda_0,\lambda_2,M)\,\},\qquad(17)V={i∈Supp(β^​)c:∣⟨r^,Xi​⟩∣>c(λ0​,λ2​,M)},(17)

where c(λ0,λ2,M)=2λ0λ2c(\lambda_0,\lambda_2,M)=2\sqrt{\lambda_0\lambda_2}c(λ0​,λ2​,M)=2λ0​λ2​​ if λ0/λ2≤M\sqrt{\lambda_0/\lambda_2}\le Mλ0​/λ2​​≤M and c(λ0,λ2,M)=λ0/M+λ2Mc(\lambda_0,\lambda_2,M)=\lambda_0/M+\lambda_2Mc(λ0​,λ2​,M)=λ0​/M+λ2​M if λ0/λ2>M\sqrt{\lambda_0/\lambda_2}>Mλ0​/λ2​​>M.

The proposition turns the optimality check of Algorithm 2 into a test on correlations with the residual. It is what makes V=∅V=\emptysetV=∅ usable in the dual-bound analysis.

Formalization Note The unit-norm assumption on the columns of XXX is the standing assumption of Section 3 of the paper; the unit norm of yyy is not needed here and is not assumed. β^\hat\betaβ^​ is an arbitrary vector: the identity does not use that it comes from Algorithm 2.

Preamble
import Mathlib
import Definitions.Def_L0BnB_DualGap_Setup
Formal statement
namespace L0BnB.DualGap

/-- Proposition 3, p. 12: with `r̂ = y − Xβ̂` and unit-norm columns of `X`,
`V = {i ∈ Supp(β̂)ᶜ | |⟨r̂, Xᵢ⟩| > c(λ₀, λ₂, M)}` (17). -/
theorem proposition_3 {n p : ℕ} (X : Matrix (Fin n) (Fin p) ℝ) (y : Fin n → ℝ)
    (lam0 lam2 M : ℝ) (hlam0 : 0 < lam0) (hlam2 : 0 < lam2) (hM : 0 < M)
    (hX : ∀ i, ∑ r, X r i ^ 2 = 1) (βhat : Fin p → ℝ) :
    Vset X y lam0 lam2 M βhat =
      Finset.univ.filter
        (fun i => βhat i = 0 ∧ cThr lam0 lam2 M < |colInner X (resid X y βhat) i|) := by sorry

end L0BnB.DualGap
Source
Hazimeh, Mazumder, Saab, Sparse Regression at Scale: Branch-and-Bound rooted in First-Order Optimization, arXiv:2004.06152v2, p. 12, Proposition 3, (17); proof p. 30; standing assumption p. 9
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 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