AI Paper Atlas

Research map / Learning theory and algorithms / Sample complexity

Adaptive Greedy Algorithm for Moderately Large Dimensions in Kernel Conditional Density Estimation

Minh-Lien Jeanne Nguyen, Claire Lacour, Vincent Rivoirard

JMLR 2022 · Learning theory and algorithms › Sample complexity · approach: Density and property estimation · 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
increase threshold constant a (ablation)RevDir CDRodeo · Example (a), n=200 000absolute errorMixed
increase decay factor beta (ablation)RevDir CDRodeo · Example (a), d1=3, n=100 000running time: beta >= 0.9Hurt
decrease decay factor beta (larger step) (ablation)RevDir CDRodeo · Example (a), d1=3, n=100 000absolute errorHurt
add irrelevant variables up to dimension 6 (ablation)RevDir CDRodeo · Example (b)estimation qualityNo clear effect
add irrelevant variables beyond dimension 10 (ablation)RevDir CDRodeo · Example (b)estimation quality: dimension 13Hurt
add relevant variables (no sparsity) (ablation)RevDir CDRodeo · Example (a)estimation quality: dimension 5Hurt
use discontinuous design (uniform inputs) instead of Gaussian inputsRevDir CDRodeo · Example (c) vs (b)estimation accuracyHurt
divide by estimated f_X instead of true f_XCDRodeo · Examples (a)-(c), n=100 000estimation accuracyImproved
add reverse step to direct CDRodeo (ablation)CDRodeo · Examples (a)-(c)estimation accuracyImproved

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