AI Paper Atlas

Research map / Learning theory and algorithms / Sample complexity

A Theory of the Risk for Optimization with Relaxation and its Application to Support Vector Machines

Marco C. Campi, Simone Garatti

JMLR 2021 · Learning theory and algorithms › Sample complexity · approach: Sample complexity of ranking and selection

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 the new two-sided complexity-based bound instead of the prior boundrisk certificate · N=1000, beta=1e-5, c=40, r=100upper bound on risk: 0.419 -> 0.2Improved
use the new bound instead of the prior boundrisk certificate · N=1000, beta=1e-5, c=320, r=250upper bound on risk: 1 -> 0.65Improved
use the new bound instead of the prior boundrisk certificate · N=2000, beta=1e-5, c=40, r=100upper bound on risk: 0.288 -> 0.10Improved
use the new bound instead of the prior boundrisk certificate · N=2000, beta=1e-5, c=10, r=0upper bound on risk: 0.04 -> 0.018Improved
increase sample size from N=1000 to N=2000new risk bound · beta=1e-5, c=40, r=100upper bound on risk: 0.2 -> 0.10Improved
decrease regularization penalty rho from 1 toward (3/5)^14SVR · noisy sinc, sigma=1complexity s* and riskHurt
choose Gaussian kernel width sigma=1 over 10^-2, 10^-1, 10, 1e1 rangeSVR · noisy sinc cost vs s*/Ncost at given complexityImproved
use smaller sigma (finer basis functions)SVR · noisy sinccost and riskMixed
select rho=(3/5)^9 via cost-risk plotSVR · noisy sinc, N=2000risk interval: [0.032, 0.08]Improved
validate Theorem 1 by Monte Carlo over 200 resamplesrisk bounds · noisy sinc, N=2000, beta=1e-4containment of true risk: 200 of 200 insideImproved

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