Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Theorem 4, p. 13 — ½‖v_{k+1}‖² − ½‖v_k‖² ≤ −(α − 1)hC_k(f(x⋆) − f(x_{k+1})) − hC_k(k + 1)(f(x_k) − f(x_{k+1}))

Proved
HessianDamping.IPAHD.half_sq_vK_succ_sub_le

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

hessian-dampinginertial-algorithmlyapunovp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let H\mathcal HH be a real Hilbert space and f:H→Rf:\mathcal H\to\mathbb Rf:H→R a convex C1\mathcal C^1C1 function with a minimizer x⋆x^\starx⋆. Let α≥1\alpha\ge1α≥1, h>0h>0h>0, s=h2s=h^2s=h2, let (βk)(\beta_k)(βk​), (bk)(b_k)(bk​) be non-negative sequences, and let (xk)(x_k)(xk​) be generated by (IPAHD). With Ck=βk+1+k(βk+1−βk)−bkhkC_k=\beta_{k+1}+k(\beta_{k+1}-\beta_k)-b_khkCk​=βk+1​+k(βk+1​−βk​)−bk​hk and vk=(α−1)(xk−x⋆)+k(xk−xk−1+βkh∇f(xk))v_k=(\alpha-1)(x_k-x^\star)+k(x_k-x_{k-1}+\beta_kh\nabla f(x_k))vk​=(α−1)(xk​−x⋆)+k(xk​−xk−1​+βk​h∇f(xk​)), if k≥1k\ge1k≥1 and the growth condition (G2dis)(\mathcal G_2^{\rm dis})(G2dis​) holds at kkk, i.e. bkhk−βk+1−k(βk+1−βk)>0b_khk-\beta_{k+1}-k(\beta_{k+1}-\beta_k)>0bk​hk−βk+1​−k(βk+1​−βk​)>0, then

12∥vk+1∥2−12∥vk∥2≤−(α−1)hCk(f(x⋆)−f(xk+1))−hCk(k+1)(f(xk)−f(xk+1)).\tfrac12\|v_{k+1}\|^2-\tfrac12\|v_k\|^2\le-(\alpha-1)hC_k\big(f(x^\star)-f(x_{k+1})\big)-hC_k(k+1)\big(f(x_k)-f(x_{k+1})\big).21​∥vk+1​∥2−21​∥vk​∥2≤−(α−1)hCk​(f(x⋆)−f(xk+1​))−hCk​(k+1)(f(xk​)−f(xk+1​)).

This is the key estimate on the quadratic part of the Lyapunov energy EkE_kEk​ of Theorem 4.

Formalization Note. The minimizer is carried as a point x⋆x^\starx⋆ with f(x⋆)≤f(y)f(x^\star)\le f(y)f(x⋆)≤f(y) for all yyy (the paper's "Given x⋆∈argminfx^\star\in\mathrm{argmin} fx⋆∈argminf"). h>0h>0h>0 and βk,bk≥0\beta_k,b_k\ge0βk​,bk​≥0 are implicit on the page and added; βk+1≥0\beta_{k+1}\ge0βk+1​≥0 is what makes the dropped coefficient of ∥∇f(xk+1)∥2\|\nabla f(x_{k+1})\|^2∥∇f(xk+1​)∥2 non-positive.

Preamble
import Mathlib
import Definitions.Def_GoldenRatioVI_Shared_IsProxPoint
import Definitions.Def_HessianDamping_IPAHD_Setting
Formal statement
namespace HessianDamping.IPAHD

/-- Proof of Theorem 4, p. 13: the bound on `½‖v_{k+1}‖² - ½‖v_k‖²` under (G₂dis) at `k`. -/
theorem half_sq_vK_succ_sub_le
    {H : Type*} [NormedAddCommGroup H] [InnerProductSpace ℝ H] [CompleteSpace H]
    (f : H → ℝ) (hfconv : ConvexOn ℝ Set.univ f) (hfC1 : ContDiff ℝ 1 f)
    (xstar : H) (hxstar : ∀ y, f xstar ≤ f y)
    (α : ℝ) (hα : 1 ≤ α)
    (h s : ℝ) (hh : 0 < h) (hs : s = h ^ 2)
    (β b : ℕ → ℝ) (hβ : ∀ k, 0 ≤ β k) (hb : ∀ k, 0 ≤ b k)
    (x : ℕ → H) (hx : IsIPAHDRun f α s β b x)
    (k : ℕ) (hk : 1 ≤ k)
    (hG2 : 0 < b k * h * k - β (k + 1) - k * (β (k + 1) - β k)) :
    1 / 2 * ‖vK f α h β x xstar (k + 1)‖ ^ 2 - 1 / 2 * ‖vK f α h β x xstar k‖ ^ 2
      ≤ -(α - 1) * h * CK h β b k * (f xstar - f (x (k + 1)))
        - h * CK h β b k * ((k : ℝ) + 1) * (f (x k) - f (x (k + 1))) := by sorry

end HessianDamping.IPAHD
Source
Attouch, Chbani, Fadili, Riahi, First-order optimization algorithms via inertial systems with Hessian driven damping, arXiv:1907.10536v2, §3.1.1, proof of Theorem 4, p. 13
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