Research map / Learning theory and algorithms / Prediction and decision losses
Censored Quantile Regression Neural Networks for Distribution-Free Survival Analysis
Tim Pearce, Jong-Hyeon Jeong, yichen jia, Jun Zhu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace sequential per-quantile NN grid with single-NN CQRNN | censored quantile regression NN · Norm linear, 9 quantiles | training speed up: 14.7$\times$ | Improved |
| replace sequential grid with CQRNN, 19 quantiles | censored quantile regression NN · Norm same | training speed up: 34.3$\times$ | Improved |
| replace sequential grid with CQRNN | censored quantile regression NN · type 1 datasets | parameter saving: 8.3$\times$ (9 quantiles), 16.2$\times$ (19 quantiles) | Improved |
| replace sequential grid with CQRNN | censored quantile regression NN · 14 type 1 datasets | TQMSE: 8/14 CQRNN better, 3/14 seq. grid better | Mixed |
| CQRNN vs sequential grid | censored quantile regression NN · Norm linear | TQMSE: 2.779 -> 0.088 | Improved |
| CQRNN vs excluding censored points | censored quantile regression NN · Norm linear | TQMSE: 0.566 -> 0.088 | Improved |
| CQRNN vs DeepQuantReg | censored quantile regression NN · Norm linear | TQMSE: 1.172 -> 0.088 | Improved |
| CQRNN vs LogNorm MLE | censored quantile regression NN · Norm linear | TQMSE: 0.184 -> 0.088 | Improved |
| CQRNN vs sequential grid on Norm uniform | censored quantile regression NN · Norm uniform | TQMSE: 0.175 $\pm$ 0.033 | Hurt |
| CQRNN vs sequential grid | censored quantile regression NN · type 3 datasets | C-index and CensDCal: 4/7 C-index, 7/7 CensDCal | Improved |
| larger quantile grid M in {9,19,39} (ablation) | CQRNN · type 1 datasets | TQMSE | Mixed |
| set y* to 1.2 times the max training y (ablation) | CQRNN · benchmarks | accuracy of quantiles | No clear effect |
Closest papers by meaning
- Deep Copula-Based Survival Analysis for Dependent Censoring with Identifiability Guarantees (AAAI 2024 · similarity 0.86)
- Boosting Methods for Interval-censored Data with Regression and Classification (ICLR 2025 · similarity 0.85)
- Orthogonal Survival Learners for Estimating Heterogeneous Treatment Effects from Time-to-Event Data (NeurIPS 2025 · similarity 0.85)
- Simulation-Based Inference with Quantile Regression (ICML 2024 · similarity 0.84)
- Relaxed Quantile Regression: Prediction Intervals for Asymmetric Noise (ICML 2024 · similarity 0.84)
- Censor Dependent Variational Inference (ICML 2025 · similarity 0.84)
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.
