Theorem 5 — Concavity in the number of segments
ProvedFBMSP.theorem5_segment_concavityFor any nonnegative integrable prediction law and independent integrable mean-zero MHR error, the marginal revenue from another segment is nonincreasing:
This target retains arbitrary prediction distributions. The pricing structure is essential; concavity is not assumed for arbitrary dynamic-program block scores.
import Definitions.Def_FBMSP_Model
namespace FBMSP
theorem theorem5_segment_concavity (ν η : MeasureTheory.Measure ℝ)
(hν : ValidPredictions ν) (hη : ValidNoise η) (hMHR : HasMHR η)
(k : ℕ) (hk : 1 ≤ k) :
optimalRevenue ν η (k + 2) - optimalRevenue ν η (k + 1) ≤
optimalRevenue ν η (k + 1) - optimalRevenue ν η k := by sorry
end FBMSPRead-back
What the Lean code literally says, in plain math · gpt-6-astra
Let be probability measures on for which the identity function is integrable, with for -almost every and . Write , , , and . Empty menus are included and yield zero; these quantities optimize over all nonnegative real prices and do not assume any maximizing price or menu has been chosen. Assume also that is absolutely continuous with respect to Lebesgue measure and that is convex and is concave on , meaning for and ; zero survival values outside are allowed. For every natural , the assertion is . In particular the smallest allowed compares the gain from two to three allowed prices with the gain from one to two; the case is not asserted. There is no assumption that is nonnegative under the two laws, and no second-moment or bounded-support assumption.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.