Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§2.1, pp. 4–5 — under nesting (1) and loss monotonicity (2), the risk λ ↦ R(λ) is nonincreasing on Λ

Proved
RiskControl.UCB.risk_antitone

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

monotonicityp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1prediction-setsrisk-control

Let PPP be a measure on X×Y\mathcal X \times \mathcal YX×Y, let Λ⊆R‾\Lambda \subseteq \overline{\mathbb R}Λ⊆R, and let {Tλ}\{\mathcal T_\lambda\}{Tλ​} be a family of set-valued predictors Tλ:X→2Z\mathcal T_\lambda : \mathcal X \to 2^{\mathcal Z}Tλ​:X→2Z that is nested on Λ\LambdaΛ,

λ1<λ2  ⟹  Tλ1(x)⊆Tλ2(x)(λ1,λ2∈Λ, x∈X),(1)\lambda_1 < \lambda_2 \implies \mathcal T_{\lambda_1}(x) \subseteq \mathcal T_{\lambda_2}(x) \quad (\lambda_1, \lambda_2 \in \Lambda,\ x \in \mathcal X), \tag{1}λ1​<λ2​⟹Tλ1​​(x)⊆Tλ2​​(x)(λ1​,λ2​∈Λ, x∈X),(1)

and let LLL be a loss on sets satisfying

S⊆S′  ⟹  L(y,S)≥L(y,S′).(2)S \subseteq S' \implies L(y, S) \ge L(y, S'). \tag{2}S⊆S′⟹L(y,S)≥L(y,S′).(2)

Suppose that (x,y)↦L(y,Tλ(x))(x, y) \mapsto L(y, \mathcal T_\lambda(x))(x,y)↦L(y,Tλ​(x)) is PPP-integrable for every λ∈Λ\lambda \in \Lambdaλ∈Λ. Then the risk R(λ)=E[L(Y,Tλ(X))]R(\lambda) = \mathbb E[L(Y, \mathcal T_\lambda(X))]R(λ)=E[L(Y,Tλ​(X))] is nonincreasing on Λ\LambdaΛ:

λ1,λ2∈Λ, λ1≤λ2  ⟹  R(λ2)≤R(λ1).\lambda_1, \lambda_2 \in \Lambda,\ \lambda_1 \le \lambda_2 \implies R(\lambda_2) \le R(\lambda_1).λ1​,λ2​∈Λ, λ1​≤λ2​⟹R(λ2​)≤R(λ1​).

This monotonicity of the risk is what lets the paper turn a pointwise confidence bound into a guarantee for the data-driven choice λ^\hat\lambdaλ^ (p. 5); Theorem A.1 assumes it, and Theorem 1 obtains it from (1) and (2).

Formalization Note The paper names this property on p. 5 without a separate statement. The integrability hypothesis is needed because the Bochner integral returns 000 for a non-integrable integrand.

Preamble
import Mathlib
import Definitions.Def_RiskControl_UCB_Setting

open MeasureTheory
Formal statement
namespace RiskControl.UCB

/-- §2.1, pp. 4–5: under the nesting property (1) of `{𝒯_λ}` and the monotonicity (2) of the
loss, the risk `λ ↦ R(λ)` is nonincreasing on `Λ`. -/
theorem risk_antitone {𝒳 𝒴 𝒵 : Type*} [MeasurableSpace 𝒳] [MeasurableSpace 𝒴]
    (P : Measure (𝒳 × 𝒴)) (Λ : Set EReal)
    (T : EReal → 𝒳 → Set 𝒵)
    (hnest : ∀ l₁ ∈ Λ, ∀ l₂ ∈ Λ, l₁ < l₂ → ∀ x, T l₁ x ⊆ T l₂ x)
    (L : 𝒴 → Set 𝒵 → ℝ)
    (hLmono : ∀ y S S', S ⊆ S' → L y S' ≤ L y S)
    (hint : ∀ l ∈ Λ, Integrable (fun p : 𝒳 × 𝒴 => L p.2 (T l p.1)) P) :
    AntitoneOn (risk P L T) Λ := by sorry

end RiskControl.UCB
Source
Bates, Angelopoulos, Lei, Malik & Jordan, arXiv:2101.02703v3, §2.1, (1), (2) and the risk R(𝒯), p. 4; monotonicity of the risk named on p. 5, paragraph after Theorem 1
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 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