Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Eq. (14), one bit — the error probability of any estimate of a binary parameter is at least the minimum Bayes risk

Proved
FastCLO.LowerBound.eq_14

by mikedeng1 · Oct 9, 2026 · Mathlib 0df444a (Lean v4.33.1)

bayes-risklower-boundsp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Let a binary parameter b∈{0,1}b \in \{0,1\}b∈{0,1} and an observation D\mathcal DD taking finitely many values ω\omegaω have joint probabilities w0(ω)=P~(b=0,D=ω)w_0(\omega) = \tilde{\mathbb P}(b = 0, \mathcal D = \omega)w0​(ω)=P~(b=0,D=ω) and w1(ω)=P~(b=1,D=ω)w_1(\omega) = \tilde{\mathbb P}(b = 1, \mathcal D = \omega)w1​(ω)=P~(b=1,D=ω), and let b^(ω)∈{0,1}\hat b(\omega) \in \{0,1\}b^(ω)∈{0,1} be any estimate of bbb. Then

∑ωmin⁡{w0(ω),w1(ω)}  ≤  ∑ω: b^(ω)=1w0(ω)+∑ω: b^(ω)=0w1(ω).\sum_\omega \min\{w_0(\omega), w_1(\omega)\} \;\le\; \sum_{\omega:\ \hat b(\omega) = 1} w_0(\omega) + \sum_{\omega:\ \hat b(\omega) = 0} w_1(\omega).ω∑​min{w0​(ω),w1​(ω)}≤ω: b^(ω)=1∑​w0​(ω)+ω: b^(ω)=0∑​w1​(ω).

The left side is EP~[min⁡{P~(b=1∣D),1−P~(b=1∣D)}]\mathbb E_{\tilde{\mathbb P}}\big[\min\{\tilde{\mathbb P}(b=1\mid\mathcal D), 1-\tilde{\mathbb P}(b=1\mid\mathcal D)\}\big]EP~​[min{P~(b=1∣D),1−P~(b=1∣D)}], the minimum Bayes risk; the right side is P~(b^≠b)\tilde{\mathbb P}(\hat b \ne b)P~(b^=b).

This is Eq. (14) of the paper for a single coordinate bib_ibi​; summing over coordinates and multiplying by ζρ(Z)/η\zeta\rho(\mathcal Z)/\etaζρ(Z)/η gives the display. It is the step that turns the regret of an arbitrary algorithm into a quantity that depends only on the data distribution.

Formalization Note The expectations are written as finite sums over the observation space, which covers the finitely supported data of the paper's construction.

Preamble
import Mathlib
Formal statement
namespace FastCLO.LowerBound

/-- Eq. (14) for one bit (Hu, Kallus, Mao, arXiv:2011.03030v3, proof of Theorem 3, p. 21; reused in
the proof of Theorem 7, p. 30): the probability that an estimate `b̂(D)` of a binary parameter `b`
is wrong is at least the minimum Bayes risk `E[min{P(b = 1 | D), 1 − P(b = 1 | D)}]`.

Formalization Note: the data take finitely many values `ω`; `w0 ω = P(b = 0, D = ω)` and
`w1 ω = P(b = 1, D = ω)`. Then `E[min posterior] = Σ_ω min(w0 ω, w1 ω)` and
`P(b̂ ≠ b) = Σ_ω (w0 ω if b̂(ω) = 1 else w1 ω)`. -/
theorem eq_14 {Ω : Type*} [Fintype Ω] (w0 w1 : Ω → ℝ) (h0 : ∀ ω, 0 ≤ w0 ω) (h1 : ∀ ω, 0 ≤ w1 ω)
    (bhat : Ω → Bool) :
    ∑ ω, min (w0 ω) (w1 ω) ≤ ∑ ω, (if bhat ω then w0 ω else w1 ω) := by sorry

end FastCLO.LowerBound
Source
Hu, Kallus, Mao, Fast Rates for Contextual Linear Optimization, arXiv:2011.03030v3, proof of Theorem 3 (A.2), Eq. (14), p. 21; reused in the proof of Theorem 7, p. 30
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