Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Lemma 3, Eq. (15): one-step expected cost inequality for modified AIM

Proved
AIMInventory.UnknownBound.eq_15

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

inventoryonline-convex-optimizationp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1stochastic-gradient

Fix a period t≥1t\ge1t≥1 for which the modified AIM projection interval contains the newsvendor quantity: yNV≤tδ/2y^{NV}\le t^{\delta/2}yNV≤tδ/2. Write ϵt=1/(max⁡{b,h}t)\epsilon_t=1/(\max\{b,h\}\sqrt t)ϵt​=1/(max{b,h}t​), let y^tδ\hat y^\delta_ty^​tδ​ be the target and Ht(y^tδ)H_t(\hat y^\delta_t)Ht​(y^​tδ​) its sales-based update signal. Then

E[Q(y^tδ)−Q(yNV)]≤E∣y^tδ−yNV∣22ϵt−E∣y^t+1δ−yNV∣22ϵt+ϵt2E∣Ht(y^tδ)∣2.\mathbb E[Q(\hat y^\delta_t)-Q(y^{NV})]\le\frac{\mathbb E|\hat y^\delta_t-y^{NV}|^2}{2\epsilon_t}-\frac{\mathbb E|\hat y^\delta_{t+1}-y^{NV}|^2}{2\epsilon_t}+\frac{\epsilon_t}{2}\mathbb E|H_t(\hat y^\delta_t)|^2.E[Q(y^​tδ​)−Q(yNV)]≤2ϵt​E∣y^​tδ​−yNV∣2​−2ϵt​E∣y^​t+1δ​−yNV∣2​+2ϵt​​E∣Ht​(y^​tδ​)∣2.

This is the one-step comparison used for the rounds after the growing projection interval reaches yNVy^{NV}yNV.

Formalization Note The paper applies Appendix A, Eq. (15), to the varying projection intervals of Theorem 7. The comparator is the fixed newsvendor quantity on the rounds where it belongs to the current interval. The common hypotheses are h,b>0h,b>0h,b>0, 0<δ<1/20<\delta<1/20<δ<1/2, finite-mean positive-mean demand, and y^1δ∈[0,1]\hat y^\delta_1\in[0,1]y^​1δ​∈[0,1].

Preamble
import Definitions.Def_AIMInventory_UnknownBound_AIM

open MeasureTheory
Formal statement
namespace AIMInventory.UnknownBound

/-- Equation (15) of Appendix A applied to a round of the modified AIM run
whose projection interval contains the newsvendor comparator. -/
theorem eq_15 (P : CappedBaseStock.DemandLaw) (δ b h y₁ : ℝ) (t : ℕ)
    (hδ : δ ∈ Set.Ioo (0 : ℝ) (1 / 2)) (hb : 0 < b) (hh : 0 < h)
    (hy₁ : y₁ ∈ Set.Icc (0 : ℝ) 1) (ht : 1 ≤ t)
    (hcomp : newsvendor P h b ≤ radius δ t) :
    (∫ d : CappedBaseStock.DemandPath,
        (cost P h b (modAimTarget δ b h y₁ d t) - cost P h b (newsvendor P h b))
          ∂CappedBaseStock.demandPathLaw P) ≤
      (∫ d : CappedBaseStock.DemandPath,
          (modAimTarget δ b h y₁ d t - newsvendor P h b) ^ 2
            ∂CappedBaseStock.demandPathLaw P) / (2 * stepSize b h t) -
      (∫ d : CappedBaseStock.DemandPath,
          (modAimTarget δ b h y₁ d (t + 1) - newsvendor P h b) ^ 2
            ∂CappedBaseStock.demandPathLaw P) / (2 * stepSize b h t) +
      stepSize b h t / 2 *
        (∫ d : CappedBaseStock.DemandPath,
          (signal h b (modAimTarget δ b h y₁ d t) (d (t - 1))) ^ 2
            ∂CappedBaseStock.demandPathLaw P) := by sorry

end AIMInventory.UnknownBound
Source
Huh & Rusmevichientong, A Non-Parametric Asymptotic Analysis of Inventory Planning with Censored Demand, Math. Oper. Res. (2009), author's manuscript, proof of Theorem 7, p. 12, applying proof of Lemma 3 Eq. (15), p. 16; https://doi.org/10.1287/moor.1080.0355
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