Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Theorem 6, predictor step — from N(1/4)\mathcal N(1/4)N(1/4) into N(1/2)\mathcal N(1/2)N(1/2) with θk+1/θk≤1−8−1/4/n+1\theta^{k+1}/\theta^k\le1-8^{-1/4}/\sqrt{n+1}θk+1/θk≤1−8−1/4/n+1​

Proved
SelfDualLP.Complexity.predictor_step

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

interior-point-methodslinear-programmingp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1predictor-corrector

Work under the choice (7). Let zk=(yk,xk,τk,θk,sk,κk)∈N(1/4)z^k=(y^k,x^k,\tau^k,\theta^k,s^k,\kappa^k)\in\mathcal N(1/4)zk=(yk,xk,τk,θk,sk,κk)∈N(1/4), let ddd solve (11)–(12) at zkz^kzk with γ=0\gamma=0γ=0, and let

αˉ=max⁡{α: zk+αd∈N(1/2)}\bar\alpha=\max\{\alpha:\ z^k+\alpha d\in\mathcal N(1/2)\}αˉ=max{α: zk+αd∈N(1/2)}

be the step size (13), assumed to exist. Then zk+1=zk+αˉd∈N(1/2)z^{k+1}=z^k+\bar\alpha d\in\mathcal N(1/2)zk+1=zk+αˉd∈N(1/2) and

θk+1θk=(xk+1)Tsk+1+τk+1κk+1(xk)Tsk+τkκk≤1−8−1/4n+1.\frac{\theta^{k+1}}{\theta^k}=\frac{(x^{k+1})^Ts^{k+1}+\tau^{k+1}\kappa^{k+1}}{(x^k)^Ts^k+\tau^k\kappa^k}\le 1-\frac{8^{-1/4}}{\sqrt{n+1}} .θkθk+1​=(xk)Tsk+τkκk(xk+1)Tsk+1+τk+1κk+1​≤1−n+1​8−1/4​.

Each predictor step therefore reduces the gap (and θ\thetaθ) by a factor depending only on nnn.

Formalization Note The hypothesis that the maximum in (13) is attained is the paper's own presumption ("max"); the membership zk+1∈N(1/2)z^{k+1}\in\mathcal N(1/2)zk+1∈N(1/2) is part of the claim as printed. Both ratios are stated, together with their equality.

Preamble
import Mathlib
import Definitions.Def_SelfDualLP_Complexity_HLP
import Definitions.Def_SelfDualLP_Complexity_Neighborhood
import Definitions.Def_SelfDualLP_Complexity_PCSequence

open Matrix
Formal statement
namespace SelfDualLP.Complexity

/-- Proof of Theorem 6, predictor-step claim (p. 62), under (7). Let `zᵏ ∈ 𝒩(1/4)`, let `d`
solve (11)–(12) at `zᵏ` with `γ = 0`, and let `ᾱ = max {α : zᵏ + αd ∈ 𝒩(1/2)}` (13) (assumed
attained). Then `zᵏ⁺¹ = zᵏ + ᾱd ∈ 𝒩(1/2)` and
`θᵏ⁺¹/θᵏ = ((xᵏ⁺¹)ᵀsᵏ⁺¹ + τᵏ⁺¹κᵏ⁺¹)/((xᵏ)ᵀsᵏ + τᵏκᵏ) ≤ 1 − 8^{−1/4}/√(n + 1)`. -/
theorem predictor_step {m n : ℕ} (A : Matrix (Fin m) (Fin n) ℝ) (b : Fin m → ℝ)
    (c : Fin n → ℝ) (z d : HLPPoint m n) (αbar : ℝ)
    (hz : Nbhd A b c (1 / 4) z) (hd : IsPCDirection A b c z d 0)
    (hα : IsGreatest {a : ℝ | Nbhd A b c (1 / 2) (z.move d a)} αbar) :
    Nbhd A b c (1 / 2) (z.move d αbar) ∧
    (z.move d αbar).θ / z.θ = gap (z.move d αbar) / gap z ∧
    gap (z.move d αbar) / gap z ≤ 1 - (8 : ℝ) ^ (-(1 / 4 : ℝ)) / Real.sqrt ((n : ℝ) + 1) := by sorry

end SelfDualLP.Complexity
Source
Ye, Todd, Mizuno, An O(√nL)-Iteration Homogeneous and Self-Dual Linear Programming Algorithm, Math. Oper. Res. 19(1) (1994), p. 62, proof of Theorem 6, predictor-step claim; (11)–(13) on p. 60
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