AI Paper Atlas

Research map / Learning theory and algorithms / Sample complexity

Nonparametric Estimation of Non-Crossing Quantile Regression Process with Deep ReQU Neural Networks

Guohao Shen, Yuling Jiao, Yuanyuan Lin, Joel L. Horowitz, Jian Huang

JMLR 2024 · Learning theory and algorithms › Sample complexity · approach: Minimax estimation rates

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 ReQU activation instead of ReLU (ablation)DQRP · Wave n=512 tau=0.5L1 test error: 0.124 -> 0.104Improved
use ReQU activation instead of ReLU (ablation)DQRP · Triangle n=512 tau=0.5L1 test error: 0.134 -> 0.170Hurt
use ReQU activation instead of ReLU (ablation)DQRP · Triangle n=2048 tau=0.5L1 test error: 0.088 -> 0.097Hurt
joint quantile-process DQRP instead of kernel QRDQRP · Wave n=512 tau=0.5L1 test error: 0.440 -> 0.104Improved
joint quantile-process DQRP instead of QR ForestDQRP · Wave n=512 tau=0.5L1 test error: 0.157 -> 0.104Improved
increase training sample size from 512 to 2048DQRP · Wave tau=0.5L1 test error: 0.104 -> 0.073Improved
increase training sample size from 512 to 2048Kernel QR · Wave tau=0.5L1 test error: 0.440 -> 0.555Hurt
single-quantile DQR instead of DQRP process estimatorDQRP · Triangle n=512 tau=0.95L1 test error: 0.295 -> 0.198Hurt
estimate the whole process jointly with DQRP instead of interpolating DQRDQRP · Wave n=2048 tau=0.05 (process)L1 test error: 0.148 -> 0.125Improved
estimate the whole process jointly instead of interpolating QR ForestDQRP · Wave n=512 tau=0.05 (process)L1 test error: 0.535 -> 0.162Improved
use Algorithm 2 (fresh random quantile levels each minibatch) instead of Algorithm 1DQRP · Wave n=512 (qualitative)fit qualityImproved

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