AI Paper Atlas

Research map / Learning theory and algorithms / Matrix and tensor methods

Sparse SVM with Hard-Margin Loss: a Newton-Augmented Lagrangian Method in Reduced Dimensions

Penghe Zhang, Naihua Xiu, Hou-Duo Qi

JMLR 2025 · Learning theory and algorithms › Matrix and tensor methods · approach: Nonconvex low-rank and sparse optimization · 2 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 hard-margin loss with l0 cardinality constraint (iPAL) instead of l2-only hard-margin ADMM0/1iPAL vs ADMM0/1 · bre (n>m)accuracy (%): 70.93 -> 79.35Improved
use iPAL instead of ADMM0/1iPAL · col (n>m)time (sec): 2.407e+0 -> 3.817e-3Improved
use iPAL instead of ADMM0/1iPAL · chr (n<m)accuracy (%): 54.32 -> 73.29Improved
use iPAL instead of ADMM0/1iPAL · qar (n<m)accuracy (%): 77.77 -> 89.03Improved
increase penalty rho=lambdaiPAL · simulated m=1000, n=2000VFC convergenceImproved
increase penalty rho=lambdaiPAL · simulated m=1000, n=2000running timeMixed
increase muiPAL · simulated m=1000, n=2000convergence speedHurt
increase sparsity level s from 10 to 40iPAL · simulated, noise r 0.11-0.16accuracyNo clear effect
increase sparsity level siPAL · simulated m=1000, n=2000iteration convergenceImproved
increase noise rate rall solvers · simulated, m=1000, n=10000accuracy: r from 0.11 to 0.16Hurt
use iPAL instead of NLPSVMiPAL · real data n>maccuracy (%): 80.00 -> 93.39 on dorImproved

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