(2.2) — V_β(s) is nonincreasing in β
ProvedPrimalDualSubgrad.DA.eq_2_2convex-optimizationdual-averagingp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Fix a prox setting in a finite-dimensional normed space . If , then for every
Monotonicity in is what allows the scheme to increase its parameters .
Preamble
import Mathlib import Definitions.Def_PrimalDualSubgrad_DA_ProxSetting
Formal statement
namespace PrimalDualSubgrad.DA
/-- (2.2), p. 7: if `β₂ ≥ β₁ > 0`, then `V_{β₂}(s) ≤ V_{β₁}(s)` for every `s ∈ E*`. -/
theorem eq_2_2 {E : Type*} [NormedAddCommGroup E] [NormedSpace ℝ E] [FiniteDimensional ℝ E]
(P : ProxSetting E) (β₁ β₂ : ℝ) (hβ₁ : 0 < β₁) (hβ : β₁ ≤ β₂) (s : StrongDual ℝ E) :
V P β₂ s ≤ V P β₁ s := by sorry
end PrimalDualSubgrad.DA
Source
Nesterov, Primal-dual subgradient methods for convex problems, Math. Program. 120 (2009), §2, (2.2), p. 7
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.