AI Paper Atlas

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

Covariate-Guided Clusterwise Linear Regression for Generalization to Unseen Data

Dohyun Bu, Hyunho Kim, Jong-Seok Lee

ICLR 2026 · Learning theory and algorithms › Kernels and regression theory · approach: High-dimensional linear regression theory

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 shared codebook of K linear regressors with learned assignment instead of post-hoc nearest-neighbor assignment (MLR*)CG-CLR · 7 tabular regression benchmarkstest RMSE: BIKE [177.57, 178.91] -> [40.77, 41.71]Improved
use codebook (K regressors) instead of proxy prediction (ablation)CG-CLR · BIKEtest RMSE: [39.96, 40.86] vs [40.77, 41.71]No clear effect
use codebook prediction vs gradient boosting (CatBoost)CG-CLR · BIKEtest RMSE: [44.69, 45.29] -> [40.77, 41.71]Improved
use codebook prediction vs CatBoostCG-CLR · CONCRETEtest RMSE: [4.057, 4.393] vs [5.193, 5.537]Hurt
increase K from 2 to 3 (nested F-test) (ablation)CG-CLR · synthetic piecewise-linear dataF-test p-value: p < 0.001Improved
increase K from 3 to 4 (nested F-test) (ablation)CG-CLR · synthetic piecewise-linear dataF-test p-value: p = 0.038No clear effect
use learned covariate-guided assignment instead of axis-aligned splits (LDT)CG-CLR · synthetic oblique boundariesabsolute errorImproved

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