AI Paper Atlas

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

Orthant-Wise Passive Descent Algorithms for Training L1-Regularized Models

Jianqiao Wangni

AAAI 2018 · Learning theory and algorithms › Matrix and tensor methods · approach: Nonconvex low-rank and sparse optimization · 0 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 OPDA instead of a proximal algorithmOPDA · Figure 1 synthetic L1 problemfinal objective P(x): 34.621 -> 20.000Improved
use OPDA-FM instead of a proximal algorithmOPDA-FM · Figure 2 nonconvex synthetic problemfinal objective P(x): 2468.755 -> 177.038Improved
use OPDA-FM instead of Proximal-SVRGOPDA-FM · rcv1 with L2 regularization lambda2 from 0 to 100/NsuboptimalityImproved
decrease the strong convexity coefficient muOPDA-FM vs Proximal-SVRG · rcv1suboptimality gapImproved
increase the L1 regularization strength (4, 8, 32 times base)OPDA-FM vs Proximal-SVRG · rcv1suboptimality gapImproved
use OPDA-FM with no L2 regularization (non-strongly convex)OPDA-FM · rcv1 lambda2 = 0suboptimalityImproved
use OPDA-QN instead of Proximal-SVRGOPDA-QN · rcv1 and covtypesuboptimality vs datapasses and timeImproved
use Gaussian or previous-direction sketching instead of identity (BFGS)OPDA-QN · covtype (low-dimensional)suboptimalityImproved
use identity sketching (BFGS) instead of previous directionsOPDA-QN · rcv1 (high-dimensional)datapasses and running timeImproved
use Gaussian sketching instead of BFGS or prevOPDA-QN · rcv1 (high-dimensional)suboptimalityHurt
use OPDA-QN instead of OPDA-FMOPDA · rcv1 and covtypeconvergenceImproved
use OPDA-SGD instead of Proximal-SGDOPDA-SGD · CIFAR10 sparse CNNobjective vs datapassesImproved

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