Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 2 — VB(M)≥VPR(∞)V_{B(M)} \ge V_{PR(\infty)}VB(M)​≥VPR(∞)​ if M≤12λ0/λ2M \le \frac12\sqrt{\lambda_0/\lambda_2}M≤21​λ0​/λ2​​; VB(M)≤VPR(∞)V_{B(M)} \le V_{PR(\infty)}VB(M)​≤VPR(∞)​ if M≥λ0/λ2M \ge \sqrt{\lambda_0/\lambda_2}M≥λ0​/λ2​​ and λ2∈L(M)\lambda_2 \in \mathcal L(M)λ2​∈L(M)

Proved
L0BnB.BigMvsPR.proposition_2

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

mixed-integer-programmingp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1perspective-relaxationsparse-regression

Let X∈Rn×pX \in \mathbb R^{n\times p}X∈Rn×p, y∈Rny \in \mathbb R^ny∈Rn and λ0,λ2,M>0\lambda_0, \lambda_2, M > 0λ0​,λ2​,M>0. Let VB(M)V_{B(M)}VB(M)​ be the optimal value of the interval relaxation of the Big-M formulation (2) of ℓ0ℓ2\ell_0\ell_2ℓ0​ℓ2​-regularized least squares, and

VPR(∞)=min⁡β∈Rp 12∥y−Xβ∥22+∑i∈[p]2λ0 B(βiλ2/λ0)V_{PR(\infty)} = \min_{\beta\in\mathbb R^p}\ \tfrac12\|y - X\beta\|_2^2 + \sum_{i\in[p]} 2\lambda_0\,\mathcal B\big(\beta_i\sqrt{\lambda_2/\lambda_0}\big)VPR(∞)​=β∈Rpmin​ 21​∥y−Xβ∥22​+i∈[p]∑​2λ0​B(βi​λ2​/λ0​​)

the value (6) of the interval relaxation of the perspective formulation PR(∞)\mathrm{PR}(\infty)PR(∞), where B\mathcal BB is the reverse Huber penalty (4). Let S(λ2)\mathcal S(\lambda_2)S(λ2​) be the set of minimizers in (6) and L(M)={λ2>0:∃β∈S(λ2), ∥β∥∞≤M}\mathcal L(M) = \{\lambda_2 > 0 : \exists\beta\in\mathcal S(\lambda_2),\ \|\beta\|_\infty \le M\}L(M)={λ2​>0:∃β∈S(λ2​), ∥β∥∞​≤M} as in (9). Then:

  1. if M≤12λ0/λ2M \le \frac12\sqrt{\lambda_0/\lambda_2}M≤21​λ0​/λ2​​, then
VB(M)≥VPR(∞);(10)V_{B(M)} \ge V_{PR(\infty)};\qquad (10)VB(M)​≥VPR(∞)​;(10)
  1. if M≥λ0/λ2M \ge \sqrt{\lambda_0/\lambda_2}M≥λ0​/λ2​​ and λ2∈L(M)\lambda_2 \in \mathcal L(M)λ2​∈L(M), then
VB(M)≤VPR(∞).(11)V_{B(M)} \le V_{PR(\infty)}.\qquad (11)VB(M)​≤VPR(∞)​.(11)

Neither relaxation dominates the other in general: a tight Big-M bound makes the Big-M relaxation the stronger one, while a loose one makes it the weaker one, provided some optimal solution of PR(∞)\mathrm{PR}(\infty)PR(∞) lies in the box.

Formalization Note Optimal values are real infima (sInf), well defined because the objectives are nonnegative and the feasible sets nonempty. The paper remarks (p. 8) that the proposition applies for any M≥0M \ge 0M≥0; the statement here assumes M>0M > 0M>0, since the function ttt and the reformulation (37) divide by MMM; at M=0M = 0M=0 the hypothesis of (11) cannot hold.

Preamble
import Mathlib
import Definitions.Def_L0BnB_BigMvsPR_Penalties
import Definitions.Def_L0BnB_BigMvsPR_Relaxations
Formal statement
namespace L0BnB.BigMvsPR

/-- Proposition 2, p. 7: with `L(M)` as in (9),
(10) `V_{B(M)} ≥ V_{PR(∞)}` if `M ≤ ½ √(λ₀/λ₂)`, and
(11) `V_{B(M)} ≤ V_{PR(∞)}` if `M ≥ √(λ₀/λ₂)` and `λ₂ ∈ L(M)`. -/
theorem proposition_2 {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) :
    (M ≤ (1 / 2) * Real.sqrt (lam0 / lam2) → L0BnB.Strength.VPRinf X y lam0 lam2 ≤ L0BnB.Strength.VB X y lam0 lam2 M) ∧
      (Real.sqrt (lam0 / lam2) ≤ M → lam2 ∈ L X y lam0 M →
        L0BnB.Strength.VB X y lam0 lam2 M ≤ L0BnB.Strength.VPRinf X y lam0 lam2) := by sorry

end L0BnB.BigMvsPR
Source
Hazimeh, Mazumder, Saab, Sparse Regression at Scale: Branch-and-Bound rooted in First-Order Optimization, arXiv:2004.06152v2, p. 7, Proposition 2, (9)–(11); proof pp. 29–30
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