Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 2.2, (2.5), p. 4 — under Assumption A, ∃ γ > 0: limsup (1/N) log[1 − P(Ŝ^ε_N ⊂ 𝒮^ε)] ≤ −γ

Proved
KleywegtSAA.ExpRate.proposition_2_2

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

exponential-ratelarge-deviationsp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1sample-average-approximation

Let S\mathcal SS be a nonempty finite set, G:S×W→RG : \mathcal S \times \mathcal W \to \mathbb RG:S×W→R, and W1,W2,…W^1, W^2, \dotsW1,W2,… an i.i.d. sample of the random vector WWW such that G(x,⋅)G(x, \cdot)G(x,⋅) is measurable and E∣G(x,W)∣<∞\mathbb E|G(x,W)| < \inftyE∣G(x,W)∣<∞ for every x∈Sx \in \mathcal Sx∈S. Let g(x)=E G(x,W)g(x) = \mathbb E\, G(x,W)g(x)=EG(x,W) and g^N(x)=N−1∑n=1NG(x,Wn)\hat g_N(x) = N^{-1}\sum_{n=1}^N G(x, W^n)g^​N​(x)=N−1∑n=1N​G(x,Wn), and let Sε\mathcal S^\varepsilonSε and S^Nε\hat{\mathcal S}^\varepsilon_NS^Nε​ be the sets of ε\varepsilonε-optimal solutions of min⁡Sg\min_{\mathcal S} gminS​g and min⁡Sg^N\min_{\mathcal S} \hat g_NminS​g^​N​.

Suppose Assumption A holds: for every x∈Sx \in \mathcal Sx∈S the moment generating function of G(x,W)G(x,W)G(x,W) is finite in a neighbourhood of 000. Then for every ε≥0\varepsilon \ge 0ε≥0 there is a constant γ>0\gamma > 0γ>0 such that

lim sup⁡N→∞1Nlog⁡[1−P(S^Nε⊂Sε)]≤−γ.\limsup_{N \to \infty} \frac1N \log\Big[1 - P\big(\hat{\mathcal S}^\varepsilon_N \subset \mathcal S^\varepsilon\big)\Big] \le -\gamma.N→∞limsup​N1​log[1−P(S^Nε​⊂Sε)]≤−γ.

That is, the probability that every ε\varepsilonε-optimal solution of the sample average approximation is an ε\varepsilonε-optimal solution of the true problem tends to one exponentially fast in the sample size.

Formalization Note The standing setting of §1–§2 is carried as hypotheses: S\mathcal SS is a nonempty finite set of an arbitrary type, each G(x,⋅)G(x,\cdot)G(x,⋅) is measurable, the sample W1,W2,…W^1, W^2, \dotsW1,W2,… is an independent, identically distributed sequence of measurable maps on a probability space (Ω,P)(\Omega, P)(Ω,P) (the paper's WnW^nWn is the (n−1)(n-1)(n−1)-st term of a 000-indexed sequence, and the law of WWW is that of the first term), and E∣G(x,W)∣<∞E|G(x,W)| < \inftyE∣G(x,W)∣<∞ for every x∈Sx \in \mathcal Sx∈S. The logarithm is ENNReal.log with log⁡0=−∞\log 0 = -\inftylog0=−∞ and the limit superior is taken in the extended reals, so the statement remains correct (with value −∞-\infty−∞) when the event is sure, e.g. when Sε=S\mathcal S^\varepsilon = \mathcal SSε=S. The constant γ\gammaγ may depend on ε\varepsilonε, as on the page, where it depends on α(ε)\alpha(\varepsilon)α(ε). The term at N=0N = 0N=0 is 000 and does not affect the limit superior.

Preamble
import Mathlib
import Definitions.Def_KleywegtSAA_ExpRate_Setting

open MeasureTheory ProbabilityTheory Filter Topology
Formal statement
namespace KleywegtSAA.ExpRate

/-- Proposition 2.2 and (2.5), p. 4: under Assumption A, for every `ε ≥ 0` there is `γ > 0` with
`limsup_{N → ∞} (1/N) log[1 − P(Ŝ^ε_N ⊂ S^ε)] ≤ −γ` (with `log 0 = −∞`). -/
theorem proposition_2_2
    {X : Type*} (S : Finset X) (hS : S.Nonempty)
    {𝒲 : Type*} [MeasurableSpace 𝒲] (G : X → 𝒲 → ℝ) (hG : ∀ x ∈ S, Measurable (G x))
    {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P]
    (W : ℕ → Ω → 𝒲) (hWm : ∀ n, Measurable (W n)) (hind : iIndepFun W P)
    (hid : ∀ n, IdentDistrib (W n) (W 0) P P)
    (hint : ∀ x ∈ S, Integrable (fun ω => G x (W 0 ω)) P)
    (hA : AssumptionA S G P W) :
    ∀ ε : ℝ, 0 ≤ ε → ∃ γ : ℝ, 0 < γ ∧
      Filter.limsup (fun N : ℕ => ((1 / (N : ℝ) : ℝ) : EReal) *
          ENNReal.log (1 - P {ω | epsSet S hS (sampleObj G W N ω) ε ⊆
            epsSet S hS (trueObj G P W) ε})) atTop ≤ ((-γ : ℝ) : EReal) := by sorry

end KleywegtSAA.ExpRate
Source
Kleywegt & Shapiro, The sample average approximation method for stochastic discrete optimization, preprint (two-author version, sha256 56657748…), p. 4, Proposition 2.2 and (2.5)
Human review
  • Endorsed by Shuze Chen · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 7, 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