AI Paper Atlas

Research map / Learning theory and algorithms / Sample complexity

Error Analysis for Deep ReLU Feedforward Density-Ratio Estimation with Bregman Divergence

Siming Zheng, Guohao Shen, Yuanyuan Lin, Jian Huang

JMLR 2026 · Learning theory and algorithms › Sample complexity · approach: Statistical analysis of GAN estimators · 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 mixing chain (mTRE) instead of convolution chain (cTRE)telescoping density ratio estimation · Beta (d=5), n=5000, K=5MSE: 1.4670 -> 0.9850Improved
use mixing chain instead of convolution chaintelescoping density ratio estimation · Normal setting, n=5000, K=10MSE: 2.7076 -> 2.8330Hurt
use mixing chain instead of convolution chaintelescoping density ratio estimation · Normal setting, n=5000, K=5MSE: 2.7987 -> 2.7426No clear effect
use mixing chain instead of convolution chain, chain length 5telescoping density ratio estimation · MNISTANLL: 1.41 -> 1.40Improved
use mixing chain with chain length 10 instead of convolution chaintelescoping density ratio estimation · MNISTANLL: NAN failure -> 1.39Improved
increase training sample size n from 100 to 100000BDD · Model (TN)average MSE: 3.48502 -> 0.00376Improved
increase training sample size n from 100 to 100000BDD · Model (NI)average MSE: 6.32138 -> 0.17945Improved
make X heavy tailed (Student-t(2)) violating Assumption 12 instead of normalBDD · Assumption 12 check, n=100000average MSE: 0.19501 -> 1.95932Hurt
let alpha K^-1 approach 1 (0.9 to 0.9999999)TRE · d=1average MSE: 1.473×101 -> 1.542×10−7Improved
let alpha K^-1 approach 1 (0.9 to 0.9999999)TRE · d=5average MSE: 5.296×1013 -> 6.689×10−2Improved
increase test sample size m from 100 to 400 in KL divergence CIKL divergence estimate · d=5, n=5000, rho=0.4coverage probability: 0.935 -> 0.805Hurt

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