§4, p. 336 — the lexicographic program has a solution, and every solution is nondominated
ProvedSteuerChoo.Lexico.lexMin_exists_and_nondominatedlexicographic-optimizationmulti-objective-optimizationp2o-batch-p200ap2o-gran-per-chapterp2o-plan-paperp2o-v1pareto-optimality
Let (), and weights . Then:
- if is compact and nonempty, the lexicographic weighted Tchebycheff program with weights has a minimizer;
- every minimizer of the lexicographic weighted Tchebycheff program with weights is nondominated:
This makes precise the remark that if the first-stage minimization of does not yield a nondominated vector, the second-stage minimization of moves to one that is.
Formalization Note Part 1 uses compactness of , an added hypothesis (the paper assumes bounded and speaks of minimizers). Part 2 needs neither compactness nor the ideal-vector property, and is arbitrary in both parts.
Preamble
import Mathlib import Definitions.Def_SteuerChoo_Lexico_nondominated import Definitions.Def_SteuerChoo_Lexico_IsLexMin
Formal statement
namespace SteuerChoo.Lexico
theorem lexMin_exists_and_nondominated {k : ℕ} [NeZero k]
(Z : Set (Fin k → ℝ)) (zstar lam : Fin k → ℝ)
(hlam : lam ∈ stdSimplex ℝ (Fin k)) :
(IsCompact Z → Z.Nonempty → ∃ z, IsLexMin Z lam zstar z) ∧
∀ z, IsLexMin Z lam zstar z → z ∈ nondominated Z := by sorry
end SteuerChoo.Lexico
Source
Steuer and Choo, An Interactive Weighted Tchebycheff Procedure for Multiple Objective Programming, Math. Programming 26 (1983), p. 336, §4, paragraph after Theorem 4.6
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.