Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 2, p. 11 — a dependent Bernoulli process X satisfies ‖𝓛(X) − 𝓛(Y)‖ ≤ 2(2b₁ + 2b₂ + b₃) for the Poisson process Y with intensity p_(·)

Proved
ChenStein.Process.theorem_2

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

chen-stein-methoddependent-indicatorsp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1poisson-approximationpoisson-processtotal-variation

Let III be an index set and (Xα)α∈I(X_\alpha)_{\alpha\in I}(Xα​)α∈I​ Bernoulli random variables on a probability space, with pα=P(Xα=1)>0p_\alpha=P(X_\alpha=1)>0pα​=P(Xα​=1)>0 and

λ=∑α∈Ipα∈(0,∞).\lambda=\sum_{\alpha\in I}p_\alpha\in(0,\infty).λ=α∈I∑​pα​∈(0,∞).

For each α\alphaα choose Bα⊆IB_\alpha\subseteq IBα​⊆I with α∈Bα\alpha\in B_\alphaα∈Bα​, and let

b1=∑α∈I∑β∈Bαpαpβ,b2=∑α∈I∑α≠β∈BαE(XαXβ),b3=∑α∈IE∣E{Xα−pα∣σ(Xβ:β∈I−Bα)}∣.b_1=\sum_{\alpha\in I}\sum_{\beta\in B_\alpha}p_\alpha p_\beta,\qquad b_2=\sum_{\alpha\in I}\sum_{\alpha\ne\beta\in B_\alpha}E(X_\alpha X_\beta),\qquad b_3=\sum_{\alpha\in I}E\bigl|E\{X_\alpha-p_\alpha\mid\sigma(X_\beta:\beta\in I-B_\alpha)\}\bigr|.b1​=α∈I∑​β∈Bα​∑​pα​pβ​,b2​=α∈I∑​α=β∈Bα​∑​E(Xα​Xβ​),b3​=α∈I∑​E​E{Xα​−pα​∣σ(Xβ​:β∈I−Bα​)}​.

For α∈I\alpha\in Iα∈I let YαY_\alphaYα​ be Poisson with mean pαp_\alphapα​, the YαY_\alphaYα​ mutually independent. Then the dependent Bernoulli process X=(Xα)α∈I\mathbf X=(X_\alpha)_{\alpha\in I}X=(Xα​)α∈I​ and the Poisson process Y=(Yα)α∈I\mathbf Y=(Y_\alpha)_{\alpha\in I}Y=(Yα​)α∈I​ on III with intensity p(⋅)p_{(\cdot)}p(⋅)​ satisfy

∥L(X)−L(Y)∥≤2(2b1+2b2+b3),\|\mathcal L(\mathbf X)-\mathcal L(\mathbf Y)\|\le2(2b_1+2b_2+b_3),∥L(X)−L(Y)∥≤2(2b1​+2b2​+b3​),

where ∥L(X)−L(Y)∥=2sup⁡A∣P(X∈A)−P(Y∈A)∣\|\mathcal L(\mathbf X)-\mathcal L(\mathbf Y)\|=2\sup_A|P(\mathbf X\in A)-P(\mathbf Y\in A)|∥L(X)−L(Y)∥=2supA​∣P(X∈A)−P(Y∈A)∣ is the paper's total variation norm (twice the usual distance).

When b1,b2,b3b_1,b_2,b_3b1​,b2​,b3​ are small, the locations of the dependent events are approximately a Poisson process; the bound needs only the first two moments pαp_\alphapα​, pαβp_{\alpha\beta}pαβ​ when XαX_\alphaXα​ is independent of the variables outside BαB_\alphaBα​ (b3=0b_3=0b3​=0).

Formalization Note The laws are the image measures of ω↦(Xα(ω))α\omega\mapsto(X_\alpha(\omega))_\alphaω↦(Xα​(ω))α​ and ω′↦(Yα(ω′))α\omega'\mapsto(Y_\alpha(\omega'))_\alphaω′↦(Yα​(ω′))α​ on NI\mathbb N^INI with the product σ-algebra; Y\mathbf YY lives on its own probability space, since total variation compares only laws. Mutual independence of the YαY_\alphaYα​ is a hypothesis (iIndepFun). The norm is 2 * tvDist, with tvDist the supremum over measurable sets of ∣μ(A)−ν(A)∣|\mu(A)-\nu(A)|∣μ(A)−ν(A)∣. b1,b2,b3b_1,b_2,b_3b1​,b2​,b3​ are [0,∞][0,\infty][0,∞]-valued and the inequality is read in [0,∞][0,\infty][0,∞] (an infinite bbb makes the bound vacuous, as on the page). The index set is assumed countable, which loses nothing (pα>0p_\alpha>0pα​>0 and ∑pα<∞\sum p_\alpha<\infty∑pα​<∞ force it) and makes the process measurable. The Bernoulli variables are N\mathbb NN-valued and bounded by 111.

Preamble
import Mathlib
import Definitions.Def_TotalVariationDist
import Definitions.Def_ChenStein_Process_Setting
Formal statement
namespace ChenStein.Process

open MeasureTheory ProbabilityTheory
open scoped ENNReal NNReal

theorem theorem_2
    {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P]
    {I : Type*} [Countable I]
    (X : I → Ω → ℕ) (hXm : ∀ α, Measurable (X α)) (hX01 : ∀ α ω, X α ω ≤ 1)
    (hp : ∀ α, 0 < p P X α)
    (lam : ℝ) (hlam : HasSum (p P X) lam) (hlam0 : 0 < lam)
    (B : I → Set I) (hB : ∀ α, α ∈ B α)
    {Ω' : Type*} [MeasurableSpace Ω'] (P' : Measure Ω') [IsProbabilityMeasure P']
    (Y : I → Ω' → ℕ) (hYm : ∀ α, Measurable (Y α)) (hYind : iIndepFun Y P')
    (hYlaw : ∀ α, P'.map (Y α) = poissonMeasure (p P X α).toNNReal) :
    ENNReal.ofReal (2 * MarkovChainCLT.tvDist (P.map (Xproc X)) (P'.map (Xproc Y)))
      ≤ 2 * (2 * b1 P X B + 2 * b2 P X B + b3 P X B) := by sorry

end ChenStein.Process
Source
Arratia, Goldstein and Gordon, Two moments suffice for Poisson approximations: the Chen-Stein method, Ann. Probab. 17 (1989), p. 11, Theorem 2; setting of §2, p. 10
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 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