Equation (15), p. 5556 — experimental price variation grows as √t
ProvedBanKeskin.KnownSparsity.eq_15dynamic-pricingexplorationleast-squaresp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Let , where selects experimental periods and is the mean price on those periods. Under ILSX with distinct experimental prices and , for every ,
The bound supplies the deterministic amount of price variation used in the information and estimation-error estimates.
Formalization Note is computed from the fixed experimental prices and is therefore independent of the customer sample.
Preamble
import Mathlib import Definitions.Def_BanKeskin_KnownSparsity_Model open MeasureTheory
Formal statement
namespace BanKeskin.KnownSparsity
/-- Price-variation growth (15), p. 5556. -/
theorem eq_15 {d : ℕ} {S : Finset (Fin (d + 1))} {Ω : Type*}
[MeasurableSpace Ω] (P : Measure Ω) (M : Model d S Ω P)
(t : ℕ) (ht : 5 ≤ t) :
(1 / 8 : ℝ) * (M.m1 - M.m2) ^ 2 * Real.sqrt (t : ℝ) ≤ M.J t := by sorry
end BanKeskin.KnownSparsity
Source
Ban and Keskin, Personalized Dynamic Pricing with Machine Learning, Management Science 67(9) (2021), p. 5556, (13) and (15)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.