Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Proposition 5, p. 26 (corrected) — if R̂⁺_WSR < R then maxᵢ 𝒦ᵢ(R) ≥ 1/δ

Proved
RiskControl.WSR.ucb_lt_imp_max_ge

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

betting-confidence-boundp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1risk-controlupper-confidence-bound

Let n∈Nn \in \mathbb Nn∈N, δ∈R\delta \in \mathbb Rδ∈R, let ω=(L1,…,Ln)\omega = (L_1, \dots, L_n)ω=(L1​,…,Ln​) be a sample with every Lj∈[0,1]L_j \in [0,1]Lj​∈[0,1], and let R≥0R \ge 0R≥0. If R^WSR+<R\widehat R^+_{\mathrm{WSR}} < RRWSR+​<R, then

max⁡i=1,…,nKi(R)≥1δ,\max_{i=1,\dots,n} \mathcal K_i(R) \ge \frac1\delta,i=1,…,nmax​Ki​(R)≥δ1​,

that is, Ki(R)≥1/δ\mathcal K_i(R) \ge 1/\deltaKi​(R)≥1/δ for some i∈{1,…,n}i \in \{1, \dots, n\}i∈{1,…,n} (see the definition RiskControl.WSR.Process for Ki\mathcal K_iKi​ and R^WSR+\widehat R^+_{\mathrm{WSR}}RWSR+​).

Applied at the true mean R=R(λ)R = R(\lambda)R=R(λ), this places the failure event {R^WSR+<R(λ)}\{\widehat R^+_{\mathrm{WSR}} < R(\lambda)\}{RWSR+​<R(λ)} inside the event controlled by Ville's inequality.

Formalization Note The paper writes "if R^WSR+(λ)<R(λ)\widehat R^+_{\mathrm{WSR}}(\lambda) < R(\lambda)RWSR+​(λ)<R(λ), then P(max⁡i=1,…,nKi≥1/δ)P(\max_{i=1,\dots,n}\mathcal K_i \ge 1/\delta)P(maxi=1,…,n​Ki​≥1/δ)", with a stray "P(P(P("; the intended claim, stated here, is the event max⁡iKi(R)≥1/δ\max_i \mathcal K_i(R) \ge 1/\deltamaxi​Ki​(R)≥1/δ. The comparison R^WSR+<R\widehat R^+_{\mathrm{WSR}} < RRWSR+​<R is in the extended reals. The hypothesis R≥0R \ge 0R≥0 holds for the paper's R=R(λ)R = R(\lambda)R=R(λ), the mean of a nonnegative loss.

Preamble
import Mathlib
import Definitions.Def_RiskControl_WSR_Process
Formal statement
namespace RiskControl.WSR

/-- Proof of Proposition 5, p. 26 (arXiv:2101.02703v3), with the page's stray `P(` removed: for
a sample of losses in `[0, 1]` and `R ≥ 0`, if `R̂⁺_WSR < R` then `max_{i=1,…,n} 𝒦_i(R; λ) ≥ 1/δ`. -/
theorem ucb_lt_imp_max_ge (n : ℕ) (δ : ℝ) (ω : Fin n → ℝ)
    (hω : ∀ j, ω j ∈ Set.Icc (0 : ℝ) 1) (R : ℝ) (hR : 0 ≤ R) :
    wsrUCB n δ ω < (R : EReal) → ∃ i ∈ Finset.Icc 1 n, 1 / δ ≤ K n δ ω i R := by sorry

end RiskControl.WSR
Source
Bates, Angelopoulos, Lei, Malik & Jordan, arXiv:2101.02703v3, Proof of Proposition 5, p. 26, last sentence before the third display
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