Proof of Theorem 2, p. 1215 — only two competing rules in T*ₙ, viz. t(a) and t(1)
ProvedSamuelCahnProphet.IID.supE_threePoint_eq_maxiidp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1prophet-inequalitythreshold-rule
Fix reals , , and an integer with . Let be i.i.d., taking the values , and with probabilities , and . Then the best threshold rule in is or :
The paper states that "there are essentially only two competing rules in , viz. and "; the identity above is that claim. It reduces the gambler's side of the extremal example to two explicit expectations.
Formalization Note. consists of and for all , both families included. The hypothesis is the page's, and makes the law a probability measure.
Preamble
import Mathlib import Definitions.Def_SamuelCahnProphet_IID_Setting open MeasureTheory Filter Topology
Formal statement
namespace SamuelCahnProphet.IID
theorem supE_threePoint_eq_max (a b c : ℝ) (ha : 0 < a) (ha1 : a < 1) (hb : 0 < b)
(hc : 0 < c) (n : ℕ) [NeZero n] (hn : b + c < n) :
supE (threePoint n a b c) n =
max (EstopT (threePoint n a b c) n a) (EstopT (threePoint n a b c) n 1) := by sorry
end SamuelCahnProphet.IID
Source
Samuel-Cahn, Comparison of threshold stop rules and maximum for independent nonnegative random variables, Ann. Probab. 12 (1984), p. 1215, proof of Theorem 2 ("There are essentially only two competing rules in T*ₙ, viz. t(a) and t(1).")
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.