Proposition 11, p. 613 — nearest neighbor weights satisfy condition (1) with C = β(d, a/b)
ProvedStoneRegression.NearestNeighbor.proposition_11Let be i.i.d. in with law , let be a regular sequence of scales with constants in (7), let and let be the nearest neighbor probability weight function (8) of a coefficient row . If is a nonnegative Borel function on with , then
This is condition (1) of Stone's consistency criterion for nearest neighbor weights, with a constant that depends only on the dimension and the ratio — in particular not on nor on the distribution of .
Formalization Note The expectations are lower Lebesgue integrals in , and takes values in . The hypothesis is the section's standing assumption that the scales are regular (p. 599), which contains (7) with the constants of the statement and the measurability of the scales.
import Mathlib import Definitions.Def_StoneRegression_Criterion_Setting import Definitions.Def_StoneRegression_NearestNeighbor_Weights
namespace StoneRegression.NearestNeighbor
open MeasureTheory ProbabilityTheory Filter Topology
open scoped ENNReal NNReal
theorem proposition_11 {d : ℕ} (μ : Measure (EuclideanSpace ℝ (Fin d))) [IsProbabilityMeasure μ]
(s : ScaleSeq d) (a b : ℝ) (hs : IsRegular μ s a b)
(c : ℕ → ℕ → ℝ) (n : ℕ) (hn : 1 ≤ n) (hc : IsCoeffRow c n)
(f : EuclideanSpace ℝ (Fin d) → ℝ≥0) (hf : Measurable f) (hfi : ∫⁻ x, (f x : ℝ≥0∞) ∂μ < ∞) :
∫⁻ ω, ∑ i : Fin n, ENNReal.ofReal (StoneRegression.Criterion.wAt (nnWeights c s) n ω i) * (f (ω (i.val + 1)) : ℝ≥0∞) ∂(StoneRegression.Criterion.xLaw μ) ≤
(beta d (a / b) : ℝ≥0∞) * ∫⁻ x, (f x : ℝ≥0∞) ∂μ := by sorry
end StoneRegression.NearestNeighbor
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.