Scaled kernel ratios are strictly ordered by the state feature
ProvedPGDRevision20261113.ordered_kernel_ratiobregman-geometryconvex-optimizationproximal-gradient
Use either admissible geometry model of Theorem 5.1, with ordering feature and positive kernels . Let and . Whenever two states satisfy ,
For the interval model and ; for the real-line model and . This is the ordered-kernel comparison used by the classification certificates.
Related results. Exact profile performance (Theorem 4.1) · Geometry classification (Theorem 5.1)
Preamble
import Definitions.Def_PGDRevision20261113_Geometry open PGDRevision20261113
Formal statement
theorem PGDRevision20261113.ordered_kernel_ratio : OrderedKernelRatioClaim := by sorry
Source
Author manuscript: Exact Residual Bounds and Optimal Geometry Selection for Proximal Gradient Methods, revised reading edition, 13 November 2026, Theorem 5.1 proof, kernel facts, pages 13-14.