Proposition 4.3: with a convex denied-service cost, stochastically larger demand to come gives greatest optimal overbooking limits that are no larger
DisprovedRevenueManagement.overbooking_limits_demandIn the dynamic overbooking model with a convex denied-service cost, let be a second family of demand pmfs with stochastically larger than in every period, that is for all and . Then the greatest optimal overbooking limit of the model with demands is at most that of the model with demands , in every period . This is the book's statement with the parametrized family , stochastically increasing in , replaced by two of its members.
Retired. Retired because the demand pmfs were not required to have finite means, so the tsum expectation in valueGo collapsed to 0 for a heavy-tailed demand and the comparison reversed (accepted disproof); replaced by RevenueManagement.overbooking_limits_demand_v2 with finite-mean hypotheses.
import Definitions.Def_RevenueManagement_overbooking
namespace RevenueManagement
theorem overbooking_limits_demand (M : DynOverbooking) (hM : M.IsModel) (hc : IsConvexSeq M.c)
(f' : ℕ → ℕ → ℝ) (hf' : ∀ t, (∀ d, 0 ≤ f' t d) ∧ HasSum (f' t) 1)
(hst : ∀ t k, ∑' d, (if k ≤ d then M.f t d else 0) ≤ ∑' d, (if k ≤ d then f' t d else 0))
(t : ℕ) (ht : 1 ≤ t) (htT : t ≤ M.T) :
({ M with f := f' } : DynOverbooking).overbookingLimit t ≤ M.overbookingLimit t := by sorry
end RevenueManagement
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.