AFTD Auto-Formalizing Theoretical Domains

subsampled_prophet_opt

def

Given a measurable space Ω, a measure P on Ω, a natural number n, random variables X : Fin n → Ω → ℝ, and a subsampling rate r : ℝ, subsampled_prophet_opt P n X r is defined as the sum over all subsets s : Finset (Fin n) of subsample_prob n r s multiplied by the integral with respect to P of finset_sup_zero s (fun i => X i ω).

Statement

open MeasureTheory ProbabilityTheory Finset in
noncomputable def subsampled_prophet_opt {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω)
    (n : ℕ) (X : Fin n → Ω → ℝ) (r : ℝ) : ℝ
Defined in
arXiv:2611.15711, Section 2.1
Verified
14 Nov 2026
Axioms
none listed
Built on 2
  • finset_sup_zeroThe maximum value of a function x on a finite subset s of Fin n, defined to be 0 when s is empty.
  • subsample_probThe probability of obtaining subset s of Fin n under independent subsampling where each of the n items is included independently with…

Read back from the Lean

Given a measurable space Ω, a measure P on Ω, a natural number n, a family of functions X : Fin n → Ω → ℝ, and a real number r, subsampled_prophet_opt P n X r is defined as the sum over all subsets s of Fin n (elements of Finset (Fin n)) of the product of subsample_prob n r s and the Bochner integral with respect to P of ω ↦ finset_sup_zero s (fun i => X i ω).

Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.

Lean source view module on GitHub

open MeasureTheory ProbabilityTheory Finset in
/-- Expected offline prophet value under independent subsampling with probability r. -/
noncomputable def subsampled_prophet_opt {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω)
    (n : ℕ) (X : Fin n → Ω → ℝ) (r : ℝ) : ℝ := ∑ s : Finset (Fin n), subsample_prob n r s * ∫ ω, finset_sup_zero s (fun i => X i ω) ∂P
See it in the graphDiscuss this resultChallenge it