Research map / Probabilistic and scientific modeling / Differential equations and PDEs
SODEN: A Scalable Continuous-Time Survival Model through Ordinary Differential Equation Networks
Weijing Tang, Jiaqi Ma, Qiaozhu Mei, Ji 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 |
|---|---|---|---|
| remove PH assumption (SODEN vs SODEN-PH) (ablation) | SODEN · MIMIC | NLL: 0.436 -> 0.411 | Improved |
| remove PH assumption (SODEN vs SODEN-PH) (ablation) | SODEN · SUPPORT | NLL: 0.702 -> 0.676 | Improved |
| remove PH assumption (SODEN vs SODEN-PH) (ablation) | SODEN · METABRIC | NLL: 0.176 -> 0.149 | Improved |
| remove PH assumption (SODEN vs SODEN-PH) (ablation) | SODEN · MIMIC-SEQ | NLL: 0.520 -> 0.489 | Improved |
| use non-linear neural hazard instead of linear Cox form (SODEN-Cox to SODEN) (ablation) | SODEN · SUPPORT | NLL: 0.761 -> 0.676 | Improved |
| use non-linear neural hazard instead of linear Cox form (SODEN-Cox to SODEN) (ablation) | SODEN · METABRIC | C-index (G=1e-8): 0.642 -> 0.661 | Improved |
| use non-PH SODEN instead of SODEN-PH (ablation) | SODEN · MIMIC (G=0.4) | C-index: 0.734 -> 0.746 | Improved |
| use non-PH SODEN instead of SODEN-PH (ablation) | SODEN · SUPPORT (G=1e-8) | C-index: 0.608 -> 0.627 | Improved |
| SODEN likelihood objective vs DeepSurv partial likelihood | SODEN · SUPPORT (G=1e-8) | C-index: 0.609 -> 0.627 | Improved |
| neural network models vs linear Cox | DeepSurv · MIMIC (G=1e-8) | C-index: 0.660 -> 0.683 | Improved |
| DeepHit with ranking loss vs SODEN on calibration | DeepHit · SUPPORT (G=1e-8) | IBS: 0.191 -> 0.197 | Hurt |
| discretizing time (DeepHit) vs continuous-time SODEN on large data | DeepHit · MIMIC-SEQ (G=1e-8) | C-index: 0.688 vs 0.687 | No clear effect |
Later papers that compared with this method
1 later accepted paper reports SODEN as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- A Scalable Nonparametric Continuous-Time Survival Model through Numerical Quadrature (NeurIPS 2026 · better on 5 of 8 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- DeepHit — DeepHit: A Deep Learning Approach to Survival Analysis With Competing Risks (AAAI 2018) · better on 0 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- A Scalable Nonparametric Continuous-Time Survival Model through Numerical Quadrature (NeurIPS 2026 · similarity 0.85)
- Deep Copula-Based Survival Analysis for Dependent Censoring with Identifiability Guarantees (AAAI 2024 · similarity 0.85)
- Gradient-based Explanations for Deep Learning Survival Models (ICML 2025 · similarity 0.85)
- NeuralSurv: Deep Survival Analysis with Bayesian Uncertainty Quantification (NeurIPS 2025 · similarity 0.85)
- Learning Survival Models with Right-Censored Reporting Delays (NeurIPS 2026 · similarity 0.85)
- Orthogonal Survival Learners for Estimating Heterogeneous Treatment Effects from Time-to-Event Data (NeurIPS 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.
