Proof of Theorem 8.3.4 — t*(T*) + T* ≤ 2 t_opt
ProvedMatousekLP.Scheduling.best_T_le_two_toptapproximation-algorithmslinear-programmingp2o-batch-b23bp2o-gran-per-chapterp2o-plan-bookp2o-v1scheduling
Let be running times of jobs on machines, let be an optimal schedule with makespan , and write for the optimal value of the linear program (with when is infeasible). Let be a real number minimizing over all real , and let be an optimal solution of . Then
Combined with Lemma 8.3.3 at , this bounds the makespan of the rounded schedule by twice the optimum.
Formalization Note Minimality of is the hypothesis for every real and every optimal solution of ; values for which has no optimal solution impose no condition, matching the convention .
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): `t*(T*) + T* ≤ 2 t_opt`. Here
`t_opt` is the makespan of an optimal schedule, `T*` minimizes `t*(T) + T` over all
real `T` (with `t*(T) = ∞` when `LPR(T)` is infeasible), and `t* = t*(T*)` is the
value of an optimal solution of `LPR(T*)`. The minimality of `T*` is stated as
`t* + T* ≤ t + T` for every `T` and every optimal solution `(t, x)` of `LPR(T)`. -/
theorem best_T_le_two_topt {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)
(Tstar tstar : ℝ) (xstar : Matrix (Fin m) (Fin n) ℝ)
(hopt : LPROptimal d Tstar tstar xstar)
(hmin : ∀ (T t : ℝ) (x : Matrix (Fin m) (Fin n) ℝ),
LPROptimal d T t x → tstar + Tstar ≤ t + T) :
tstar + Tstar ≤ 2 * 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 (displayed inequality); t*(T) = ∞ convention p. 154
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.