Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3.9 (Consistency), pp. 21–22 — measurable optimal estimates converge μ-a.s. to the unique minimizer in D

Proved
DupacovaWets.Consistency.consistency_of_measurable_estimates

by mikedeng1 · Oct 4, 2026 · Mathlib 0df444a (Lean v4.33.1)

consistencyepi-convergencep2o-batch-pfp1bp2o-gran-per-chapterp2o-plan-paperp2o-v1stochastic-programming

The standing data are those of §3: Ξ\XiΞ is a Polish space with its Borel σ\sigmaσ-field, PPP a probability measure on it, (Z,F,μ)(Z,\mathcal F,\mu)(Z,F,μ) a probability space with an increasing sequence of sub-σ\sigmaσ-fields F1⊆F2⊆⋯⊆F\mathcal F^1\subseteq\mathcal F^2\subseteq\dots\subseteq\mathcal FF1⊆F2⊆⋯⊆F, Pν(⋅,ζ)P^\nu(\cdot,\zeta)Pν(⋅,ζ) random probability measures on Ξ\XiΞ, and f:Rn×Ξ→(−∞,∞]f:\mathbb R^n\times\Xi\to(-\infty,\infty]f:Rn×Ξ→(−∞,∞] an integrand with domain S×ΞS\times\XiS×Ξ; fff, SSS and the measures satisfy Assumptions 3.4 and 3.5. Write Ef(x)=∫Ξf(x,ξ) P(dξ)Ef(x)=\int_\Xi f(x,\xi)\,P(d\xi)Ef(x)=∫Ξ​f(x,ξ)P(dξ) and Eνf(x,ζ)=∫Ξf(x,ξ) Pν(dξ,ζ)E^\nu f(x,\zeta)=\int_\Xi f(x,\xi)\,P^\nu(d\xi,\zeta)Eνf(x,ζ)=∫Ξ​f(x,ξ)Pν(dξ,ζ), with the +∞+\infty+∞ convention of p. 10.

Suppose Assumptions 3.4 and 3.5 hold and there is a compact set D⊆RnD\subseteq\mathbb R^nD⊆Rn such that for every ν=1,2,…\nu=1,2,\dotsν=1,2,… the set (argmin⁡Eνf)∩D(\operatorname{argmin}E^\nu f)\cap D(argminEνf)∩D is nonempty μ\muμ-almost surely, and

{x∗}=argmin⁡Ef∩D.\{x^*\}=\operatorname{argmin}Ef\cap D .{x∗}=argminEf∩D.

Then there exist Fν\mathcal F^\nuFν-measurable selections xνx^\nuxν of argmin⁡Eνf\operatorname{argmin}E^\nu fargminEνf such that

x∗=lim⁡ν→∞xν(ζ)for μ-almost all ζ,x^*=\lim_{\nu\to\infty}x^\nu(\zeta)\quad\text{for }\mu\text{-almost all }\zeta,x∗=ν→∞lim​xν(ζ)for μ-almost all ζ,

and also inf⁡Ef=lim⁡ν→∞(inf⁡Eνf)\inf Ef=\lim_{\nu\to\infty}(\inf E^\nu f)infEf=limν→∞​(infEνf) μ\muμ-almost surely.

This is the strong consistency statement of the paper: estimators that use only the information available at stage ν\nuν and solve the estimated problem converge almost surely to the true solution, and the estimated optimal values converge to the true optimal value. It extends Wald's and Huber's consistency theorems to constrained problems with extended-real-valued, merely lower semicontinuous criteria.

Formalization Note Lean writes Rn\mathbb R^nRn as EuclideanSpace ℝ (Fin n), and the sample index is shifted: Lean's Pν k ζ and 𝔽 k are the paper's Pk+1(⋅,ζ)P^{k+1}(\cdot,\zeta)Pk+1(⋅,ζ) and Fk+1\mathcal F^{k+1}Fk+1, while P=P0P = P^0P=P0 is kept as the separate argument P. The paper's selections are maps xν:Z0→Rnx^\nu:Z_0\to\mathbb R^nxν:Z0​→Rn, measurable for the trace of Fν\mathcal F^\nuFν on a full-measure set Z0Z_0Z0​; here they are total maps xν:Z→Rnx^\nu:Z\to\mathbb R^nxν:Z→Rn, Fν\mathcal F^\nuFν-measurable on all of ZZZ, with xν(ζ)∈argmin⁡Eνf(⋅,ζ)x^\nu(\zeta)\in\operatorname{argmin}E^\nu f(\cdot,\zeta)xν(ζ)∈argminEνf(⋅,ζ) for all ν\nuν for μ\muμ-almost every ζ\zetaζ. The two forms are equivalent: a function into Rn\mathbb R^nRn that is measurable for a trace σ\sigmaσ-field extends measurably to the whole space, and Z0Z_0Z0​ has full measure. The limit of the infima is in [−∞,∞][-\infty,\infty][−∞,∞]. The paper also says that Ξ\XiΞ is the support of PPP; this hypothesis is not used in the proofs of §3 and is omitted, which makes the statement stronger.

Preamble
import Mathlib
import Definitions.Def_DupacovaWets_Consistency_expect
import Definitions.Def_DupacovaWets_Consistency_ExtendedFunctions
import Definitions.Def_DupacovaWets_Consistency_Assumptions
open MeasureTheory Filter Topology
Formal statement
namespace DupacovaWets.Consistency

/-- Dupačová–Wets (WP-86-41), pp. 21–22, Theorem 3.9 (Consistency), "In particular": under
Assumptions 3.4 and 3.5, if `D ⊆ ℝⁿ` is compact, `(argmin E^ν f) ∩ D` is nonempty `μ`-a.s. for
every `ν`, and `{x*} = argmin Ef ∩ D`, then there are `F^ν`-measurable selections `x^ν` of
`argmin E^ν f` with `x^ν(ζ) → x*` for `μ`-almost all `ζ`, and `inf E^ν f → inf Ef` `μ`-a.s.
(`Pν k`, `𝔽 k`, `x k` are the paper's `P^{k+1}`, `F^{k+1}`, `x^{k+1}`.) -/
theorem consistency_of_measurable_estimates {n : ℕ} {Ξ : Type*} [TopologicalSpace Ξ] [PolishSpace Ξ] [MeasurableSpace Ξ] [BorelSpace Ξ]
    (P : Measure Ξ) [IsProbabilityMeasure P]
    {Z : Type*} [mZ : MeasurableSpace Z] (μ : Measure Z) [IsProbabilityMeasure μ]
    (𝔽 : Filtration ℕ mZ) (Pν : ℕ → Z → Measure Ξ)
    (f : EuclideanSpace ℝ (Fin n) → Ξ → EReal) (S : Set (EuclideanSpace ℝ (Fin n)))
    (h34 : Assumption3_4 f S) (h35 : Assumption3_5 f S P μ 𝔽 Pν)
    (D : Set (EuclideanSpace ℝ (Fin n))) (hD : IsCompact D)
    (hDne : ∀ k, ∀ᵐ ζ ∂μ, (argminSet (expectFn (Pν k ζ) f) ∩ D).Nonempty)
    (xstar : EuclideanSpace ℝ (Fin n)) (hx : argminSet (expectFn P f) ∩ D = {xstar}) :
    (∃ x : ℕ → Z → EuclideanSpace ℝ (Fin n), (∀ k, Measurable[𝔽 k] (x k)) ∧
        (∀ᵐ ζ ∂μ, ∀ k, x k ζ ∈ argminSet (expectFn (Pν k ζ) f)) ∧
        (∀ᵐ ζ ∂μ, Tendsto (fun k => x k ζ) atTop (𝓝 xstar))) ∧
      ∀ᵐ ζ ∂μ, Tendsto (fun k => ⨅ y, expectFn (Pν k ζ) f y) atTop (𝓝 (⨅ y, expectFn P f y)) := by sorry

end DupacovaWets.Consistency
Source
Dupačová & Wets, IIASA Working Paper WP-86-41 (Aug. 1986), pp. 21–22, Theorem 3.9, 'In particular' conclusion
Human review
  • Endorsed by Shuze Chen · Oct 4, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 4, 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