Lemma A2 — closed monotone conjugate and attained infimum
ProvedBregmanPPA.IneqMult.lemmaA2convex-conjugatemonotone-conjugatep2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be a closed proper convex function, finite and subdifferentiable at a point of the strictly positive orthant. Its monotone conjugate is closed, proper, and convex. At every point of its effective domain, the representation as an infimum over is attained:
The result guarantees a genuine minimizing conjugate argument rather than only an infimum value.
Formalization Note Both conjugates are extended-real functions, and closedness is lower semicontinuity. The effective domain excludes ; properness excludes .
Preamble
import Mathlib import Definitions.Def_BregmanPPA_IneqMult_Program open Filter Topology InertialFB.IFB
Formal statement
namespace BregmanPPA.IneqMult
/-- Lemma A2, p. 223: closedness, convexity, properness and attainment of the
restricted conjugate's infimum representation. -/
theorem lemmaA2 {m : ℕ} (F : E m → EReal)
(hproper : IsProperFn F) (hconvex : IsConvexFn F)
(hclosed : LowerSemicontinuous F)
(hu : ∃ u : E m, u ∈ posOrthant m ∧ F u ≠ ⊤ ∧
∃ v : E m, IsSubgradient F u v) :
LowerSemicontinuous (monoConjE F) ∧
IsProperFn (monoConjE F) ∧
IsConvexFn (monoConjE F) ∧
(∀ z : E m, monoConjE F z ≠ ⊤ →
∃ w : E m, (∀ i, z i ≤ w i) ∧ monoConjE F z = conjE F w) := by sorry
end BregmanPPA.IneqMult
Source
Eckstein, Nonlinear proximal point algorithms using Bregman functions, with applications to convex programming, Math. Oper. Res. 18(1) (1993), p. 223, Lemma A2, https://doi.org/10.1287/moor.18.1.202
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.