Proof of Theorem 8.3.4 — LPR(t_opt) is feasible and t*(t_opt) ≤ t_opt
ProvedMatousekLP.Scheduling.lpr_topt_lelinear-programminglp-relaxationp2o-batch-b23bp2o-gran-per-chapterp2o-plan-bookp2o-v1scheduling
Let be running times of jobs on machines, and let be an optimal schedule, with makespan . Then:
- the linear program is feasible, and
- every optimal solution of has value
This says that the relaxation with threshold is a genuine relaxation of the scheduling problem; it is the first inequality used to compare the rounded schedule with the optimum.
Formalization Note The optimal value is expressed through optimal solutions of , not through an infimum.
Preamble
import Mathlib import Definitions.Def_MatousekLP_Scheduling_Schedule import Definitions.Def_MatousekLP_Scheduling_LPRelaxation
Formal statement
namespace MatousekLP.Scheduling
/-- Proof of Theorem 8.3.4 (Matoušek–Gärtner, p. 155): with `t_opt` the makespan of an
optimal schedule, the relaxation `LPR(t_opt)` is feasible, and every optimal solution
`(t, x)` of `LPR(t_opt)` has value `t ≤ t_opt` (that is, `t*(t_opt) ≤ t_opt`). -/
theorem lpr_topt_le {m n : ℕ} (d : Matrix (Fin m) (Fin n) ℝ)
(hd : ∀ i j, 0 < d i j) (σopt : Fin n → Fin m) (hσopt : IsOptimalSchedule d σopt) :
(∃ (t : ℝ) (x : Matrix (Fin m) (Fin n) ℝ), LPRFeasible d (makespan d σopt) t x) ∧
∀ (t : ℝ) (x : Matrix (Fin m) (Fin n) ℝ),
LPROptimal d (makespan d σopt) t x → t ≤ makespan d σopt := by sorry
end MatousekLP.Scheduling
Source
Matoušek & Gärtner, Understanding and Using Linear Programming, Springer 2007, p. 155, proof of Theorem 8.3.4
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.