Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Β§3.1, proof of Theorem 3.1, (6)–(7), p. 483 β€” x* ∈ 𝒫 iff x* ∈ dom g and g(x*) + g*(y) ≀ h(x*) + h*(y) for all y ∈ dom h*

Proved
TaoAnDCA.GlobalOpt.mem_primalSol_iff_eq7

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

dc-programmingdualityglobal-optimalityp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let g,hβˆˆΞ“0(Rn)g,h\in\Gamma_0(\mathbb R^n)g,hβˆˆΞ“0​(Rn) satisfy the standing inclusions (3), with the convention +βˆžβˆ’(+∞)=+∞+\infty-(+\infty)=+\infty+βˆžβˆ’(+∞)=+∞, and let P\mathcal PP be the solution set of the d.c. program Ξ±=inf⁑{g(x)βˆ’h(x)}\alpha=\inf\{g(x)-h(x)\}Ξ±=inf{g(x)βˆ’h(x)}. For every xβˆ—βˆˆRnx^*\in\mathbb R^nxβˆ—βˆˆRn the following are equivalent:

  1. xβˆ—βˆˆPx^*\in\mathcal Pxβˆ—βˆˆP;
  2. xβˆ—βˆˆdom⁑gx^*\in\operatorname{dom}gxβˆ—βˆˆdomg and
g(xβˆ—)βˆ’h(xβˆ—)≀hβˆ—(y)βˆ’gβˆ—(y)βˆ€y∈dom⁑hβˆ—;(6)g(x^*)-h(x^*)\le h^*(y)-g^*(y)\qquad\forall y\in\operatorname{dom}h^*;\tag{6}g(xβˆ—)βˆ’h(xβˆ—)≀hβˆ—(y)βˆ’gβˆ—(y)βˆ€y∈domhβˆ—;(6)
  1. xβˆ—βˆˆdom⁑gx^*\in\operatorname{dom}gxβˆ—βˆˆdomg and
g(xβˆ—)+gβˆ—(y)≀h(xβˆ—)+hβˆ—(y)βˆ€y∈dom⁑hβˆ—.(7)g(x^*)+g^*(y)\le h(x^*)+h^*(y)\qquad\forall y\in\operatorname{dom}h^*.\tag{7}g(xβˆ—)+gβˆ—(y)≀h(xβˆ—)+hβˆ—(y)βˆ€y∈domhβˆ—.(7)

By d.c. duality, xβˆ—x^*xβˆ— is optimal exactly when its value does not exceed any value of the dual objective; (7) is the same inequality with the terms rearranged. This is the first half of the proof of Theorem 3.1(i).

Formalization Note The statement is the conjunction of "1 iff 2" and "1 iff 3". In (6) the differences are the convention-respecting dcSub; in (7) the sums are taken in the extended reals. On the ranges involved all four terms are finite: g(xβˆ—)g(x^*)g(xβˆ—) and h(xβˆ—)h(x^*)h(xβˆ—) because xβˆ—βˆˆdom⁑gβŠ‚dom⁑hx^*\in\operatorname{dom}g\subset\operatorname{dom}hxβˆ—βˆˆdomgβŠ‚domh and hhh is proper, hβˆ—(y)h^*(y)hβˆ—(y) because y∈dom⁑hβˆ—y\in\operatorname{dom}h^*y∈domhβˆ—, and gβˆ—(y)g^*(y)gβˆ—(y) because dom⁑hβˆ—βŠ‚dom⁑gβˆ—\operatorname{dom}h^*\subset\operatorname{dom}g^*domhβˆ—βŠ‚domgβˆ— and ggg is proper. The restriction y∈dom⁑hβˆ—y\in\operatorname{dom}h^*y∈domhβˆ— is kept as on the page.

Preamble
import Mathlib
import Definitions.Def_TaoAnDCA_GlobalOpt_Setting

open TaoAnDCA.GlobalOpt
Formal statement
namespace TaoAnDCA.GlobalOpt

theorem mem_primalSol_iff_eq7 {n : β„•} (g h : EuclideanSpace ℝ (Fin n) β†’ EReal)
    (hgh : DCStanding g h) (x : EuclideanSpace ℝ (Fin n)) :
    (x ∈ primalSol g h ↔ x ∈ effDom g ∧
      βˆ€ y ∈ effDom (CondatPD.FinDim.conj h),
        dcSub g h x ≀ dcSub (CondatPD.FinDim.conj h) (CondatPD.FinDim.conj g) y) ∧
    (x ∈ primalSol g h ↔ x ∈ effDom g ∧
      βˆ€ y ∈ effDom (CondatPD.FinDim.conj h),
        g x + CondatPD.FinDim.conj g y ≀ h x + CondatPD.FinDim.conj h y) := by sorry

end TaoAnDCA.GlobalOpt
Source
Pham Dinh & Le Thi, A d.c. optimization algorithm for solving the trust-region subproblem, SIAM J. Optim. 8 (1998), p. 483, Β§3.1, proof of Theorem 3.1, displays (6) and (7)
Human review
  • Endorsed by Shuze Chen Β· Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 Β· 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
Β© 2026 Prove2Me