Lemma 2.4 — moment interpolation
ProvedTightCoreset2027.Euclidean.moment_interpolationmoment-inequalitiesp2o-batch-stoc27ap2o-gran-per-chapterp2o-plan-paperp2o-v1probability
Let be a measurable real-valued random variable on any probability space. Suppose its fourth absolute moment is integrable and strictly positive, so . Then
This moment inequality is the probability-space statement used in the projection discrepancy estimate. No boundedness or finite-support restriction is imposed on . Integrability of the lower moments follows from the fourth-moment hypothesis and the probability normalization.
Differences from the source:
- encoding: a real random variable with 0 < E|Y|^4 < infinity → an arbitrary probability measure, measurable Y, integrable fourth power, and positive fourth integral; reason: these explicitly encode the meaning of random variable and finite moment and exclude the default integral on nonintegrable inputs
Preamble
import Mathlib open MeasureTheory set_option autoImplicit false
Formal statement
namespace TightCoreset2027.Euclidean
theorem moment_interpolation {Ω : Type*} [MeasurableSpace Ω]
(μ : Measure Ω) [IsProbabilityMeasure μ] (Y : Ω → ℝ)
(hY : Measurable Y)
(hfour : Integrable (fun ω => |Y ω| ^ 4) μ)
(hpos : 0 < ∫ ω, |Y ω| ^ 4 ∂μ) :
(Real.rpow (∫ ω, |Y ω| ^ 2 ∂μ) ((3 : ℝ) / 2)) /
Real.sqrt (∫ ω, |Y ω| ^ 4 ∂μ) ≤ ∫ ω, |Y ω| ∂μ := by sorry
end TightCoreset2027.Euclidean
Source
Junqiu Hu, Lingxiao Huang, Jian Li, Baoduo Xu, STOC 2027 submission #879 (October 2026), Lemma 2.4, printed p. 18 / PDF p. 20; https://github.com/sigact-stoc/stoc2027/releases/download/papers-1/879.pdf
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.