AI Paper Atlas

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

Fast, Provably convergent IRLS Algorithm for p-norm Linear Regression

Deeksha Adil, Richard Peng, Sushant Sachdeva

NeurIPS 2019 · Learning theory and algorithms › Kernels and regression theory · approach: High-dimensional linear regression theory · 14 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 p-IRLS instead of CVXℓp-regression on random matrices · Random matrix ℓp-regressionruntime: 30--50 timesImproved
Use p-IRLS instead of CVXp-Laplacian minimization · Graph p-Laplacian minimizationruntime: 10--30 timesImproved
Use p-IRLS instead of CVXℓp-regression and graph p-Laplacian minimization · Random matrices and graphsobjective valueImproved
Use p-IRLS instead of IRLS/Homotopyp-Laplacian semi-supervised learning · Hardest benchmark problem, equal iteration counterror: orders of magnitudes smallerImproved
Increase target accuracy by reducing εp-IRLS · Random matrices and graphsiterations: linearly with log(1/ε)Hurt
Increase problem sizep-IRLS · Random matrices and graphsiterations: nearly constantNo clear effect
Increase pp-IRLS · Random matrices and graphsiterations: at most linearHurt
Add systematic padding to the IRLS weights and perform a line searchStandard IRLS for ℓp-regression · ℓp-regression, p ∈ [2,∞)convergenceImproved

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