Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 5′ — an optimal x∗x^*x∗ supported on III stays optimal for any A~\tilde AA~ within ljl_jlj​ of AAA off III, over the enlarged set U~\tilde{\mathcal U}U~

Proved
RobustRegLasso.Sparsity.theorem_5_prime

by mikedeng1 · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

lassop2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1regressionrobust-optimizationsparsity

Let a1,…,am,b∈Rna_1,\dots,a_m, b \in \mathbb R^na1​,…,am​,b∈Rn and radii ci≥0c_i \ge 0ci​≥0, and let U={(δ1,…,δm)∣∥δi∥2≤ci}\mathcal U = \{(\delta_1,\dots,\delta_m) \mid \|\delta_i\|_2 \le c_i\}U={(δ1​,…,δm​)∣∥δi​∥2​≤ci​}. Let x∗x^*x∗ be an optimal solution of the robust regression problem

min⁡x∈Rm{max⁡ΔA∈U∥b−(A+ΔA)x∥2},\min_{x\in\mathbb R^m}\Big\{\max_{\Delta A\in\mathcal U}\|b-(A+\Delta A)x\|_2\Big\},x∈Rmmin​{ΔA∈Umax​∥b−(A+ΔA)x∥2​},

and let I⊆{1,…,m}I \subseteq \{1,\dots,m\}I⊆{1,…,m} be such that xj∗=0x^*_j = 0xj∗​=0 for all j∉Ij \notin Ij∈/I. For numbers ljl_jlj​ let

U~={(δ1,…,δm)∣∥δi∥2≤ci, i∈I; ∥δj∥2≤cj+lj, j∉I}.\tilde{\mathcal U} = \{(\delta_1,\dots,\delta_m) \mid \|\delta_i\|_2 \le c_i,\ i\in I;\ \|\delta_j\|_2 \le c_j + l_j,\ j\notin I\}.U~={(δ1​,…,δm​)∣∥δi​∥2​≤ci​, i∈I; ∥δj​∥2​≤cj​+lj​, j∈/I}.

Then for every A~=(a~1,…,a~m)\tilde A = (\tilde a_1,\dots,\tilde a_m)A~=(a~1​,…,a~m​) with ∥a~j−aj∥2≤lj\|\tilde a_j - a_j\|_2 \le l_j∥a~j​−aj​∥2​≤lj​ for j∉Ij \notin Ij∈/I and a~i=ai\tilde a_i = a_ia~i​=ai​ for i∈Ii\in Ii∈I, x∗x^*x∗ is an optimal solution of

min⁡x∈Rm{max⁡ΔA∈U~∥b−(A~+ΔA)x∥2}.\min_{x\in\mathbb R^m}\Big\{\max_{\Delta A\in\tilde{\mathcal U}}\|b-(\tilde A+\Delta A)x\|_2\Big\}.x∈Rmmin​{ΔA∈U~max​∥b−(A~+ΔA)x∥2​}.

This is the main theorem of the paper's Section IV: Theorem 5 is its special case with cj=0c_j = 0cj​=0 off III, and through Theorem 5 it yields the incoherence-type sparsity criterion of Theorem 6.

Formalization Note The radius vector ccc is arbitrary (not fixed), so that Theorem 5 is a specialization. The paper does not state lj≥0l_j \ge 0lj​≥0; it is implied by ∥a~j−aj∥2≤lj\|\tilde a_j - a_j\|_2 \le l_j∥a~j​−aj​∥2​≤lj​, so lll is left unconstrained. The radii ci≥0c_i \ge 0ci​≥0 are the paper's standing convention for (2) (with a negative radius U\mathcal UU is empty and every xxx is vacuously optimal, so the statement would fail). "Optimal" means minimizing the EReal-valued supremum over all of Rm\mathbb R^mRm; existence of an optimal solution is assumed, not claimed. The paper's ∥a~j−aj∥\|\tilde a_j - a_j\|∥a~j​−aj​∥ is the ℓ2\ell^2ℓ2 norm.

Preamble
import Mathlib
import Definitions.Def_RobustRegLasso_Sparsity_Basic
Formal statement
namespace RobustRegLasso.Sparsity

theorem theorem_5_prime {n m : ℕ} (a : Fin m → EuclideanSpace ℝ (Fin n))
    (b : EuclideanSpace ℝ (Fin n)) (c l : Fin m → ℝ) (hc : ∀ i, 0 ≤ c i)
    (xStar : Fin m → ℝ) (hopt : IsRobustSolution a b (RobustRegLasso.FeatureWise.uncertaintySet c) xStar)
    (I : Finset (Fin m)) (hx : ∀ j, j ∉ I → xStar j = 0) :
    ∀ aTilde : Fin m → EuclideanSpace ℝ (Fin n),
      (∀ j, j ∉ I → ‖aTilde j - a j‖ ≤ l j) → (∀ i, i ∈ I → aTilde i = a i) →
      IsRobustSolution aTilde b (enlargedSet c l I) xStar := by sorry

end RobustRegLasso.Sparsity
Source
Xu, Caramanis, Mannor, Robust Regression and Lasso, arXiv:0811.1790v1, pp. 8–9, Theorem 5′
Human review
  • Endorsed by Shuze Chen · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 7, 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