Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3, p. 12 — Z_n(x) → Z0(x) and Z̄_n(x) → Z0(x) almost surely

Proved
EmpiricalDRO.Consistency.theorem3_consistency

by mikedeng1 · Oct 8, 2026 · Mathlib 0df444a (Lean v4.33.1)

consistencydistributionally-robust-optimizationempirical-likelihoodp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1strong-law

Let ξ1,ξ2,…\xi_1,\xi_2,\dotsξ1​,ξ2​,… be independent and identically distributed random elements of a measurable space Ξ\XiΞ, with law P0P_0P0​. Fix a decision x∈Rmx\in\mathbb R^mx∈Rm and a loss h(x;⋅):Ξ→Rh(x;\cdot):\Xi\to\mathbb Rh(x;⋅):Ξ→R, measurable, such that

0<Var0(h(x;ξ))<∞,Z0(x)=E0[h(x;ξ)]<∞.0<\mathrm{Var}_0(h(x;\xi))<\infty,\qquad Z_0(x)=E_0[h(x;\xi)]<\infty .0<Var0​(h(x;ξ))<∞,Z0​(x)=E0​[h(x;ξ)]<∞.

Let 0<α<10<\alpha<10<α<1 and let χ1,1−α2\chi^2_{1,1-\alpha}χ1,1−α2​ be the (1−α)(1-\alpha)(1−α)-quantile of the χ12\chi^2_1χ12​ distribution. For each nnn let

Z‾n(x)=min⁡w∈Un(χ1,1−α2/(2n))∑i=1nh(x;ξi)wi,Z‾n(x)=max⁡w∈Un(χ1,1−α2/(2n))∑i=1nh(x;ξi)wi,\underline Z_n(x)=\min_{w\in\mathcal U_n(\chi^2_{1,1-\alpha}/(2n))}\sum_{i=1}^n h(x;\xi_i)w_i,\qquad \overline Z_n(x)=\max_{w\in\mathcal U_n(\chi^2_{1,1-\alpha}/(2n))}\sum_{i=1}^n h(x;\xi_i)w_i,Z​n​(x)=w∈Un​(χ1,1−α2​/(2n))min​i=1∑n​h(x;ξi​)wi​,Zn​(x)=w∈Un​(χ1,1−α2​/(2n))max​i=1∑n​h(x;ξi​)wi​,

where Un(η)={w:−1n∑ilog⁡(nwi)≤η, ∑iwi=1, w≥0}\mathcal U_n(\eta)=\{w:-\frac1n\sum_i\log(nw_i)\le\eta,\ \sum_iw_i=1,\ w\ge0\}Un​(η)={w:−n1​∑i​log(nwi​)≤η, ∑i​wi​=1, w≥0} is the empirical Burg-entropy divergence ball (19). Then, almost surely,

Z‾n(x)→Z0(x)andZ‾n(x)→Z0(x)(n→∞).\underline Z_n(x)\to Z_0(x)\quad\text{and}\quad\overline Z_n(x)\to Z_0(x)\qquad(n\to\infty).Z​n​(x)→Z0​(x)andZn​(x)→Z0​(x)(n→∞).

Theorem 3 complements the coverage result Theorem 2: the empirical DRO bounds, calibrated with the χ12\chi^2_1χ12​ quantile, not only bracket Z0(x)Z_0(x)Z0​(x) with asymptotic probability 1−α1-\alpha1−α but also shrink onto it.

Formalization Note The sample ξ1,…,ξn\xi_1,\dots,\xi_nξ1​,…,ξn​ is ξ 0, …, ξ (n-1). The moment conditions are MemLp (h x ∘ ξ 0) 2 (finite mean and variance) and positive variance; the quantile q=χ1,1−α2q=\chi^2_{1,1-\alpha}q=χ1,1−α2​ is pinned by N(0,1){y:y2≤q}=1−α\mathcal N(0,1)\{y:y^2\le q\}=1-\alphaN(0,1){y:y2≤q}=1−α. The paper's Θ\ThetaΘ only says where xxx lives and is dropped. The positive variance and the value of α\alphaα are kept as printed ("the same conditions in Theorem 2") although the conclusion does not need them. Z‾n\overline Z_nZn​ is the published robustMean burg (q/2) (ball radius ρ/n\rho/nρ/n with ρ=q/2\rho=q/2ρ=q/2), and Z‾n\underline Z_nZ​n​ is robustLower burg (q/2).

Preamble
import Mathlib
import Definitions.Def_EmpiricalDRO_Consistency_Setting

open MeasureTheory ProbabilityTheory Filter Topology
Formal statement
namespace EmpiricalDRO.Consistency

/-- Theorem 3, Lam, arXiv:1605.09349v1, p. 12: under the conditions of Theorem 2 (i.i.d. data,
`0 < Var₀(h(x;ξ)) < ∞`, `Z₀(x) = E₀[h(x;ξ)]` finite), for a fixed decision `x`, both empirical
DRO values `Z_n(x)` (min) and `Z̄_n(x)` (max) over the empirical Burg ball `U_n(χ²_{1,1−α}/(2n))`
converge almost surely to `Z₀(x)`. The sample `ξ_1, …, ξ_n` is `ξ 0, …, ξ (n-1)`, and `q` is the
`(1−α)`-quantile of `χ²₁` (the law of `Y²`, `Y ∼ N(0,1)`). -/
theorem theorem3_consistency {Ω : Type*} [MeasurableSpace Ω] {P : Measure Ω}
    [IsProbabilityMeasure P] {Ξ : Type*} [MeasurableSpace Ξ]
    (ξ : ℕ → Ω → Ξ) (hξ : ∀ i, Measurable (ξ i)) (hindep : iIndepFun ξ P)
    (hident : ∀ i, IdentDistrib (ξ i) (ξ 0) P P)
    {m : ℕ} (h : EuclideanSpace ℝ (Fin m) → Ξ → ℝ) (x : EuclideanSpace ℝ (Fin m))
    (hhx : Measurable (h x))
    (hL2 : MemLp (fun ω => h x (ξ 0 ω)) 2 P)
    (hvar : 0 < variance (fun ω => h x (ξ 0 ω)) P)
    (α : ℝ) (hα0 : 0 < α) (hα1 : α < 1) (q : ℝ)
    (hq : gaussianReal 0 1 {y : ℝ | y ^ 2 ≤ q} = ENNReal.ofReal (1 - α)) :
    ∀ᵐ ω ∂P,
      Tendsto (fun n : ℕ => EmpiricalDRO.Coverage.robustLower EmpiricalDRO.Coverage.burg (q / 2) (fun i : Fin n => h x (ξ i ω))) atTop
          (𝓝 (∫ ω', h x (ξ 0 ω') ∂P)) ∧
        Tendsto (fun n : ℕ =>
            GenEmpLik.Expansion.robustMean EmpiricalDRO.Coverage.burg (q / 2) (fun i : Fin n => h x (ξ i ω))) atTop
          (𝓝 (∫ ω', h x (ξ 0 ω') ∂P)) := by sorry

end EmpiricalDRO.Consistency
Source
Lam, Recovering Best Statistical Guarantees via the Empirical DRO, arXiv:1605.09349v1, p. 12, Theorem 3 (conditions of Theorem 2, p. 11; (26)–(27), p. 11); proof p. 34
Human review
  • Endorsed by Shuze Chen · Oct 9, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 9, 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