Remark after Theorem 4.1, p. 483 — a quasiconvex hemivariate function has at most one minimum point
ProvedTsengBCD.Hemivariate.atMostOneMin_of_quasiconvex_hemivariatehemivariatep2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1quasiconvexunique-minimizer
Let be a function on a real vector space that is quasiconvex,
and hemivariate (not constant on any nondegenerate line segment contained in ). Then has at most one minimum point: if and both minimize over and , then .
Applied to the section , this is the page's remark that if is quasiconvex and hemivariate in , then has at most one minimum in ; it links Assumption B2 to the uniqueness hypothesis of Theorem 4.1(c).
Formalization Note. A minimum point is a point of attaining the infimum, so the case (where every point attains the infimum) is excluded by .
Preamble
import Mathlib import Definitions.Def_TsengBCD_Hemivariate_Setting
Formal statement
namespace TsengBCD.Hemivariate
open Filter Topology
theorem atMostOneMin_of_quasiconvex_hemivariate {V : Type*} [AddCommGroup V] [Module ℝ V]
(h : V → EReal) (hqc : IsQuasiconvex h) (hhv : IsHemivariate h) :
∀ a b : V, h a ≠ ⊤ → (∀ c, h a ≤ h c) → (∀ c, h b ≤ h c) → a = b := by sorry
end TsengBCD.Hemivariate
Source
Tseng, Convergence of a block coordinate descent method for nondifferentiable minimization, J. Optim. Theory Appl. 109 (2001), p. 483, remark after Theorem 4.1
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.