AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

Data-Driven Performance Guarantees for Classical and Learned Optimizers

Rajiv Sambharya, Bartolomeo Stellato

JMLR 2025 · Learning theory and algorithms › Kernels and regression theory · approach: Information-theoretic generalization bounds · 9 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use probabilistic data-driven bounds instead of worst-case bounds for classical optimizersOSQP and SCS fixed-point iterations · Image deblurring and robust Kalman filteringquantile bound on fixed-point residualImproved
increase number of samples N from 10 to 100 to 1000 (ablation)sample convergence bound · Image deblurring and robust Kalman filteringquantile bound tightness: N=10, N=100, N=1000Improved
optimize PAC-Bayes bound with ALISTA, TiLISTA or GLISTA instead of running ISTA/FISTAlearned ISTA variants · Sparse coding (m=256, n=512, K=10)NMSE after 10 iterationsImproved
use LISTA with its very large number of weightsPAC-Bayes training · Sparse coding (m=256, n=512, K=10)NMSE boundHurt
learn warm starts with PAC-Bayes guarantees instead of cold start or nearest-neighbor warm startL2WS gradient descent · Unconstrained QP (n=20)fixed-point residual boundImproved
take the better of worst-case guarantee and probabilistic boundsample convergence bound · fixed-point residual metricrisk boundImproved
add penalty toward target regularizer value B^target in the training objectivePAC-Bayes training problem · learned optimizersbound qualityImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy