paper-with-me

홈 › Papers

When Rates Are Geometric: Rate-Certificate Transfer for Contact Splittings in Optimization

2026-07-26 · George A Kevrekidis arxiv

Discrete optimization algorithms are often analyzed through continuous-time limiting ODEs, but a convergence certificate for the ODE is not automatically one for the discrete algorithm. We develop contact Hamiltonian systems as a setting where the transfer can be made precise. A contact Hamiltonian $H$ on $J^1(\mathbb{R}^n)$ obeys the intrinsic decay identity $\dot H = -H\,\partial_s H$, so an augmented energy $\mathcal{E}$ built from $H$, together with the conformal rate $\partial_s H$, is a continuous-time rate certificate whenever $\mathcal{E}$ controls the objective gap. Our main theorem states, under three named and independently checkable hypotheses, that an order-$r$ contact splitting with step $h$ transfers this certificate over the finite horizon set by backward error analysis. The discrete decay envelope is governed by the modified conformal factor up to $O(h^r)$ perturbations plus a backward-error shadowing defect, and the mechanism is inherited exactly because the modified Hamiltonian is itself a contact Hamiltonian. Quadratic heavy ball is a fully solvable example: its projected dissipative-leapfrog spectrum agrees with established conformal-symplectic optimization theory, while the augmented contact Hamiltonian yields a sharp objective-to-certificate comparison that verifies the transfer hypotheses. For strongly convex objectives with state-dependent damping, an explicit Bregman-type Lyapunov certificate instead transfers by an auxiliary-shadowing corollary. The decomposition $H=K+V+D$ into kinetic, objective-encoding potential, and dissipation terms serves as a design template, with a catalogue of closed-form sub-flows including contact-specific damping families. Numerical experiments confirm the predicted conformal-factor tracking orders and show competitive performance on ill-conditioned benchmarks and deep-learning tasks.

📄 PDF Abstract BibTeX arXiv:2607.23642

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Restricted Dynamic Geometric Complexity: Certificates for Structured Preconditioning

2026-07-08 · Zavier Li arxiv

Optimization geometrodynamics views optimizer state as evolving geometry. Its full positive-definite quadratic benchmark gives the least affine-invariant deformation needed to reduce condition number when arbitrary metri…

Transfer of Safety Controllers Through Learning Deep Inverse Dynamics Model

2024-05-22 · Alireza Nadali, Ashutosh Trivedi, Majid Zamani

Control barrier certificates have proven effective in formally guaranteeing the safety of the control systems. However, designing a control barrier certificate is a time-consuming and computationally expensive endeavor t…

Transfer Learning

VeRecycle: Reclaiming Guarantees from Probabilistic Certificates for Stochastic Dynamical Systems after Change

2025-05-20 · Sterre Lutz, Matthijs T. J. Spaan, Anna Lukina

Autonomous systems operating in the real world encounter a range of uncertainties. Probabilistic neural Lyapunov certification is a powerful approach to proving safety of nonlinear stochastic dynamical systems. When face…

On the Tightness of Semidefinite Relaxations for Certifying Robustness to Adversarial Examples

2020-06-11 · NeurIPS 2020 12 · Richard Y. Zhang

The robustness of a neural network to adversarial examples can be provably certified by solving a convex relaxation. If the relaxation is loose, however, then the resulting certificate can be too conservative to be pract…

Adversarial Attack

Concave Certificates: Geometric Framework for Distributionally Robust Risk and Complexity Analysis

2026-01-04 · Hong T. M. Chu arxiv

Distributionally Robust (DR) optimization aims to certify worst-case risk within a Wasserstein uncertainty set. Current certifications typically rely either on global Lipschitz bounds, which are often conservative, or on…