Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Corollary 2 — Moment bound

Proved
FBMSP.corollary2_moment_bound

by mikeh · Oct 30, 2026 · Mathlib 0df444a (Lean v4.33.1)

market-segmentationoperations-researchpricing

Assume nonnegative integrable predictions with finite second moment, independent integrable mean-zero error, and nonnegative realized valuations almost surely. If prediction variance is σ2\sigma^2σ2 and σ≥0\sigma\ge0σ≥0, then for every k≥1k\ge1k≥1,

R∞−Rk≤22σ(E[M]+σ)k.R_\infty-R_k\le2\sqrt{\frac{2\sigma(\mathbb E[M]+\sigma)}{k}}.R∞​−Rk​≤2k2σ(E[M]+σ)​​.

Author-confirmed correction. The factor two was confirmed on October 30, 2026. This result uses the same nonnegative realized-valuation setting as corrected Theorem 3. Minimizing the proof's expression 2jσ/k+(E[M]+σ)/j2j\sigma/k+(\mathbb E[M]+\sigma)/j2jσ/k+(E[M]+σ)/j yields this constant. The formal proof remains open and must handle conditional truncation, boundary atoms, and the zero-variance case.

Preamble
import Definitions.Def_FBMSP_Model
open MeasureTheory
Formal statement
namespace FBMSP

theorem corollary2_moment_bound (ν η : Measure ℝ)
    (hν : ValidPredictions ν) (hη : ValidNoise η)
    (hV : NonnegativeValuations ν η)
    (hMoment : Integrable (fun x : ℝ => x ^ 2) ν)
    (σ : ℝ) (hσ : 0 ≤ σ) (hVar : predictionVariance ν = σ ^ 2)
    (k : ℕ) (hk : 1 ≤ k) :
    segmentationLoss ν η k ≤
      2 * Real.sqrt (2 * σ * (predictionMean ν + σ) / (k : ℝ)) := by sorry

end FBMSP
Source
Titing Cui and Michael L. Hamilton, Optimal Feature-Based Market Segmentation and Pricing, October 22, 2025 manuscript (60 PDF pages); Operations Research 74(3), 1354–1370 (2026). https://doi.org/10.1287/opre.2022.0490 ; https://ssrn.com/abstract=4151103 . Corollary 2, p.23; proof p.46 (ec12). Factor two confirmed by the author on October 30, 2026; the nonnegative realized-valuation condition is inherited from corrected Theorem 3.
Read-back

What the Lean code literally says, in plain math · gpt-6-astra

For every pair of probability measures ν\nuν and η\etaη on R\mathbb RR, assume that x↦xx\mapsto xx↦x is integrable with respect to ν\nuν, that x≥0x\ge0x≥0 for ν\nuν-almost every xxx, that e↦ee\mapsto ee↦e is integrable with respect to η\etaη, that ∫e dη(e)=0\int e\,d\eta(e)=0∫edη(e)=0, and that for ν\nuν-almost every xxx, the inequality x+e≥0x+e\ge0x+e≥0 holds for η\etaη-almost every eee. Assume additionally that x↦x2x\mapsto x^2x↦x2 is integrable with respect to ν\nuν. Write μ=∫x dν(x)\mu=\int x\,d\nu(x)μ=∫xdν(x). For every real number σ≥0\sigma\ge0σ≥0 satisfying ∫(x−μ)2 dν(x)=σ2\int(x-\mu)^2\,d\nu(x)=\sigma^2∫(x−μ)2dν(x)=σ2, and every natural number kkk with 1≤k1\le k1≤k, let S(t)S(t)S(t) be the real number obtained from η([t,∞))\eta([t,\infty))η([t,∞)), let Pk\mathcal P_kPk​ be the family of all finite sets P⊆[0,∞)P\subseteq[0,\infty)P⊆[0,∞) with ∣P∣≤k|P|\le k∣P∣≤k, and define GP(x)=max⁡({0}∪{pS(p−x):p∈P})G_P(x)=\max\bigl(\{0\}\cup\{pS(p-x):p\in P\}\bigr)GP​(x)=max({0}∪{pS(p−x):p∈P}) for every P∈PkP\in\mathcal P_kP∈Pk​ and every real xxx. The asserted inequality is ∫sup⁡p∈[0,∞)(pS(p−x)) dν(x)−sup⁡P∈Pk∫GP(x) dν(x) ≤ 22σ(μ+σ)k\displaystyle \int \sup_{p\in[0,\infty)}\bigl(pS(p-x)\bigr)\,d\nu(x)-\sup_{P\in\mathcal P_k}\int G_P(x)\,d\nu(x)\ \le\ 2\sqrt{\frac{2\sigma(\mu+\sigma)}{k}}∫p∈[0,∞)sup​(pS(p−x))dν(x)−P∈Pk​sup​∫GP​(x)dν(x) ≤ 2k2σ(μ+σ)​​, with kkk interpreted as a real number in the quotient. The first supremum optimizes a nonnegative price separately for each xxx, while each menu supremum chooses one fixed finite price set used for all xxx. The interval defining SSS is closed at its lower endpoint and therefore counts any mass at the posted-price threshold. Empty menus are allowed and yield G∅=0G_{\varnothing}=0G∅​=0; the zero-revenue option in the maximum does not consume a menu element, and fewer than kkk distinct prices may be used. The conclusion does not assert that an optimal price or menu is attained. Zero variance is included: σ=0\sigma=0σ=0 makes the stated upper bound 000. The case k=1k=1k=1 is included and k=0k=0k=0 is excluded, so division by zero does not arise. The square root is the total real square-root function, which returns 000 on negative inputs; here its argument is nonnegative because μ≥0\mu\ge0μ≥0, σ≥0\sigma\ge0σ≥0, and k≥1k\ge1k≥1. No second-moment assumption on η\etaη, nor any density, continuity, or hazard-rate assumption on either measure, is imposed. All displayed integrals are real Bochner integrals, whose total definition gives 000 for a nonintegrable integrand. All displayed suprema are real suprema, whose total definition gives 000 for an empty or unbounded-above set; the optimization sets here are nonempty because price 000 and the empty menu are available. Conversion of a measure value to a real sends +∞+\infty+∞ to 000, but the probability-measure hypotheses make every survival value finite.

Human review
  • Endorsed by Shuze Chen · Oct 30, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikeh · Oct 30, 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