(2.17) — ‖x^ag_k − x_k‖² ≤ Γ_k Σ_{τ≤k} ((λ_τ − β_τ)²/(Γ_τα_τ)) ‖∇Ψ(x^md_τ)‖²
ProvedNonconvexAG.Smooth.eq_2_17accelerated-gradientnonconvex-optimizationp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be generated by Algorithm 1 from with step sizes , (), , , using an arbitrary map , and let be as in (2.6). Then for every ,
Combined with (2.15), this controls the correction term of the one-step estimate by past gradient norms.
Formalization Note The statement is the outer ends of the chain (2.17). It is stated for any map g; no smoothness is used.
Preamble
import Mathlib import Definitions.Def_ConvexOptAlg_SmoothGD_Defs import Definitions.Def_NonconvexAG_Smooth_AGRun
Formal statement
namespace NonconvexAG.Smooth
/-- (2.17), p. 6 (outer ends of the chain). -/
theorem eq_2_17 {n : ℕ} (g : E n → E n) (α β lam : ℕ → ℝ) (hstep : AGStepsizes α β lam) (x0 : E n) :
∀ k, 1 ≤ k →
‖xagSeq g α β lam x0 k - xSeq g α β lam x0 k‖ ^ 2 ≤
Gamma α k * ∑ τ ∈ Finset.Icc 1 k,
(lam τ - β τ) ^ 2 / (Gamma α τ * α τ) * ‖g (xmdSeq g α β lam x0 τ)‖ ^ 2 := by sorry
end NonconvexAG.Smooth
Source
Ghadimi & Lan, Accelerated Gradient Methods for Nonconvex Nonlinear and Stochastic Programming, arXiv:1310.3787v1, p. 6, (2.17)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.