Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Survival Distributions with Individually Calibrated Asymmetric Laplace Distribution
Deming Sheng, Ricardo Henao
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 |
|---|---|---|---|
| add calibration loss L_Cal to the ALD likelihood (pre-calibration) | ALD survival model · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 39.3% better vs 0.0% worse | Improved |
| add calibration loss L_Cal to the ALD likelihood (pre-calibration) | CQRNN-style quantile loss L_Cqr · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 41.1% better vs 1.8% worse | Improved |
| add quantile regression loss L_Cqr to the ALD likelihood (pre-calibration) | ALD survival model · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 26.8% better vs 21.4% worse | Mixed |
| use L_Cal instead of L_Cqr as the additional loss (pre-calibration) (ablation) | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 64.3% wins vs. 0% losses | Improved |
| use L_Cal instead of X-CAL as additional loss (pre-calibration) (ablation) | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 44.6% better vs 5.4% worse | Improved |
| apply L_Cal as post-calibration instead of L_Cqr post-calibration (ablation) | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 76.8% better vs 3.6% worse | Improved |
| use L_Cal post-calibration instead of CSD post-calibration (ablation) | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 71.4% better vs 19.6% worse | Improved |
| use L_Cal post-calibration instead of CiPOT post-calibration (ablation) | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 55.4% better vs 5.4% worse | Improved |
| calibrate post hoc instead of pre-calibration with L_Cal | ICALD · 21 survival datasets, calibration metrics | % significantly better vs worse comparisons: 14.3% better vs 7.1% worse, 78.6% same | No clear effect |
| ICALD post-calibrated L_Cal compared with DeepSurv on IBS | ICALD · 21 survival datasets | IBS: 21 significantly better and 0 worse | Improved |
| ICALD post-calibrated L_Cal compared with DSM (Weibull) | ICALD · 21 survival datasets, 7 metrics | % significantly better: 73.6% of all comparisons (103 out of 140) | Improved |
| ICALD post-calibrated L_Cal compared with ALD on individual calibration | ICALD · 21 survival datasets | individual calibration: 9 significantly better and 0 worse | Improved |
Closest papers by meaning
- Learning Survival Distributions with the Asymmetric Laplace Distribution (ICML 2025 · similarity 0.92)
- Toward Conditional Distribution Calibration in Survival Prediction (NeurIPS 2024 · similarity 0.89)
- Conformalized Survival Distributions: A Generic Post-Process to Increase Calibration (ICML 2024 · similarity 0.87)
- Deep Copula-Based Survival Analysis for Dependent Censoring with Identifiability Guarantees (AAAI 2024 · similarity 0.85)
- Survival Analysis via Density Estimation (ICML 2025 · similarity 0.85)
- Censor Dependent Variational Inference (ICML 2025 · similarity 0.85)
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.
