Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 1: zero-setup and full-cost guarantees, tightness of 2\sqrt22​

Proved
LostSalesRQ.main_theorem

by Ming Hu · Nov 17, 2026 · Mathlib 0df444a (Lean v4.33.1)

approximation-guaranteeinventorylost-salesoperations-research

This is Theorem 1 of the paper, in full. Let λ,L,h,p,K≥0\lambda,L,h,p,K\ge0λ,L,h,p,K≥0 be the demand rate, lead time, holding cost rate, lost-sales penalty and fixed ordering cost of the continuous-review lost-sales system with unit Poisson demand, let C(π)C(\pi)C(π) be the long-run average cost of an admissible policy and OPT\mathrm{OPT}OPT the infimum over all admissible policies.

  1. Full-cost guarantee. For every nonnegative (λ,L,h,p,K)(\lambda,L,h,p,K)(λ,L,h,p,K) and every demand environment of rate λ\lambdaλ,
inf⁡π∈ΠrQC(π)≤2+62 OPT.\inf_{\pi\in\Pi_{\rm rQ}}C(\pi)\le\frac{\sqrt2+\sqrt6}{2}\,\mathrm{OPT}.π∈ΠrQ​inf​C(π)≤22​+6​​OPT.
  1. Zero-setup guarantee. When K=0K=0K=0, inf⁡π∈ΠSC(π)≤2 OPT\inf_{\pi\in\Pi_{\rm S}}C(\pi)\le\sqrt2\,\mathrm{OPT}infπ∈ΠS​​C(π)≤2​OPT on every demand environment.
  2. The algorithm at K=0K=0K=0. For h,λ>0h,\lambda>0h,λ>0, with μ=λL\mu=\lambda Lμ=λL and a=pλ/ha=p\lambda/ha=pλ/h, let S⋆S^\starS⋆ be any integer minimizing Lj(a)L_j(a)Lj​(a). On every demand environment a policy selected by the rule of Corollary 1 exists, and every such policy is an integer base-stock policy (possibly level 000) whose cost equals inf⁡π∈ΠSC(π)\inf_{\pi\in\Pi_{\rm S}}C(\pi)infπ∈ΠS​​C(π).
  3. Tightness. In normalized parameters,
sup⁡μ,a>0Uμ(a)OPT0(μ,a)=2.\sup_{\mu,a>0}\frac{U_\mu(a)}{\mathrm{OPT}_0(\mu,a)}=\sqrt2.μ,a>0sup​OPT0​(μ,a)Uμ​(a)​=2​.
  1. The ratio is well defined. For μ,a>0\mu,a>0μ,a>0, OPT0(μ,a)≥Lμ(a)≥min⁡{1,a}>0\mathrm{OPT}_0(\mu,a)\ge\mathcal L_\mu(a)\ge\min\{1,a\}>0OPT0​(μ,a)≥Lμ​(a)≥min{1,a}>0.

The constant (2+6)/2=1.93185…(\sqrt2+\sqrt6)/2=1.93185\ldots(2​+6​)/2=1.93185… is the root R>1R>1R>1 of R2=2R+1R^2=\sqrt2R+1R2=2​R+1. The theorem gives the first constant-factor guarantee valid at every finite parameter vector for this model with positive lead time and arbitrary fixed ordering costs.

Formalization Note Policy costs are evaluated on an arbitrary demand environment SSS (a probability space with a filtration carrying the Poisson demand count), while OPT\mathrm{OPT}OPT is the infimum over all environments and all admissible policies on them. Costs and OPT\mathrm{OPT}OPT are in [0,∞][0,\infty][0,∞], so the inequalities are meaningful even when OPT=0\mathrm{OPT}=0OPT=0 (for instance λ=0\lambda=0λ=0, p=0p=0p=0 or h=0h=0h=0), where the infima need not be attained. The ratio in item 4 is division in [0,∞][0,\infty][0,∞] of Uμ(a)U_\mu(a)Uμ​(a) by OPT0(μ,a)\mathrm{OPT}_0(\mu,a)OPT0​(μ,a).

Preamble
import Definitions.Def_LostSalesRQ_model
import Definitions.Def_LostSalesRQ_lower_curve
Formal statement
namespace LostSalesRQ

theorem main_theorem :
    (∀ lam L h p K : ℝ, 0 ≤ lam → 0 ≤ L → 0 ≤ h → 0 ≤ p → 0 ≤ K →
      ∀ S : DemandSpace lam,
        rQFamilyCost S L h p K ≤
          ENNReal.ofReal ((Real.sqrt 2 + Real.sqrt 6) / 2) * OPT lam L h p K) ∧
    (∀ lam L h p : ℝ, 0 ≤ lam → 0 ≤ L → 0 ≤ h → 0 ≤ p →
      ∀ S : DemandSpace lam,
        baseStockFamilyCost S L h p ≤ ENNReal.ofReal (Real.sqrt 2) * OPT lam L h p 0) ∧
    (∀ lam L h p : ℝ, 0 < lam → 0 ≤ L → 0 < h → 0 ≤ p →
      ∀ S : DemandSpace lam, ∀ Sstar : ℕ,
        (∀ j : ℕ, baseStockCost (lam * L) (p * lam / h) Sstar ≤
          baseStockCost (lam * L) (p * lam / h) j) →
        (∃ π : Policy S, IsSelectedPolicy L (lam * L) (p * lam / h) 0 Sstar π) ∧
        ∀ π : Policy S, IsSelectedPolicy L (lam * L) (p * lam / h) 0 Sstar π →
          π ∈ baseStockFamily S L ∧ π.avgCost L h p 0 = baseStockFamilyCost S L h p) ∧
    (⨆ (μ : ℝ) (a : ℝ) (_ : 0 < μ ∧ 0 < a),
        ENNReal.ofReal (minBaseStockCost μ a) / OPTk μ a 0) = ENNReal.ofReal (Real.sqrt 2) ∧
    (∀ μ a : ℝ, 0 < μ → 0 < a →
      ENNReal.ofReal (lowerCurve μ a) ≤ OPTk μ a 0 ∧ min 1 a ≤ lowerCurve μ a ∧
        0 < min 1 a) := by sorry

end LostSalesRQ
Source
M. Hu, "A 1.932 Guarantee for (r,Q) Policies in Continuous-Review Lost-Sales Inventory Systems", arXiv:2610.34576, 2026; Theorem 1, eqs. (5)-(7) (p. 8).
Human review
  • Endorsed by Shuze Chen · Nov 18, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Ming Hu · Nov 18, 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