Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

(2.21) along the RSAG run — Ψ(x^md_k) − [(1−α_k)Ψ(x^ag_{k−1}) + α_kΨ(x)] ≤ α_k⟨∇Ψ(x^md_k), x_{k−1} − x⟩

Proved
NonconvexAG.Stoch.eq_2_21

by mikedeng1 · Oct 9, 2026 · Mathlib 0df444a (Lean v4.33.1)

accelerated-gradientconvex-optimizationp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Let Ψ:Rn→R\Psi:\mathbb R^n\to\mathbb RΨ:Rn→R be convex and differentiable with LΨL_\PsiLΨ​-Lipschitz gradient. Run the RSAG method (Algorithm 3) from x0x_0x0​ with step sizes α1=1\alpha_1=1α1​=1, αk∈(0,1)\alpha_k\in(0,1)αk​∈(0,1) (k≥2k\ge2k≥2), βk,λk>0\beta_k,\lambda_k>0βk​,λk​>0, any oracle and any noise sequence. Then for every sample point, every k≥1k\ge1k≥1 and every x∈Rnx\in\mathbb R^nx∈Rn,

Ψ(xkmd)−[(1−αk)Ψ(xk−1ag)+αkΨ(x)]≤αk⟨∇Ψ(xkmd),xk−1−x⟩.\Psi(x^{md}_k)-\big[(1-\alpha_k)\Psi(x^{ag}_{k-1})+\alpha_k\Psi(x)\big]\le\alpha_k\langle\nabla\Psi(x^{md}_k),x_{k-1}-x\rangle .Ψ(xkmd​)−[(1−αk​)Ψ(xk−1ag​)+αk​Ψ(x)]≤αk​⟨∇Ψ(xkmd​),xk−1​−x⟩.

It is the convexity step of the analysis: the middle point is the convex combination (2.2) of xk−1agx^{ag}_{k-1}xk−1ag​ and xk−1x_{k-1}xk−1​, so the gradient inequality at xkmdx^{md}_kxkmd​ compares it with that combination.

Formalization Note The statement is pathwise and uses only (2.2), which Algorithm 3 shares with Algorithm 1. It is the outer ends of the paper's chain (2.21).

Preamble
import Mathlib
import Definitions.Def_ConvexOptAlg_SmoothGD_Defs
import Definitions.Def_GhadimiLan_RSG_Model
import Definitions.Def_NonconvexAG_Stoch_RSAG
open MeasureTheory ProbabilityTheory
open scoped InnerProductSpace
Formal statement
namespace NonconvexAG.Stoch

open GhadimiLan.RSG (E)

/-- (2.21), p. 7 (outer ends of the chain), along the RSAG run: for convex `Ψ`, every sample
point `ω`, every `k ≥ 1` and every `x ∈ ℝⁿ`,
`Ψ(x^md_k) − [(1 − αₖ)Ψ(x^ag_{k−1}) + αₖΨ(x)] ≤ αₖ⟨∇Ψ(x^md_k), x_{k−1} − x⟩`. -/
theorem eq_2_21 {n : ℕ} (Ψ : E n → ℝ) (g : E n → E n) (LΨ : ℝ)
    (hΨ : ConvexOptAlg.SmoothGD.IsBetaSmooth Ψ g LΨ) (hconv : ConvexOn ℝ Set.univ Ψ)
    {Ω Ξ : Type*} (G : E n → Ξ → E n) (ξ : ℕ → Ω → Ξ)
    (α β lam : ℕ → ℝ) (hstep : RSAGStepsizes α β lam) (x0 : E n) :
    ∀ k, 1 ≤ k → ∀ ω : Ω, ∀ x : E n,
      Ψ (xmdSeq G α β lam x0 ξ k ω)
          - ((1 - α k) * Ψ (xagSeq G α β lam x0 ξ (k - 1) ω) + α k * Ψ x) ≤
        α k * ⟪g (xmdSeq G α β lam x0 ξ k ω), xSeq G α β lam x0 ξ (k - 1) ω - x⟫_ℝ := by sorry

end NonconvexAG.Stoch
Source
Ghadimi & Lan, Accelerated Gradient Methods for Nonconvex Nonlinear and Stochastic Programming, arXiv:1310.3787v1, p. 7, (2.21); used on p. 17 in the proof of Theorem 3 b)
Human review
  • Endorsed by Shuze Chen · Oct 9, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 9, 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