Research map / Time series / Causal and generative time series
Temporal Phenotyping using Deep Predictive Clustering of Disease Progression
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use predictive clustering with outcome-conditioned centroids rather than the strongest listed baseline | AC-TPC · UKCF | Purity: 0.751$\pm$0.01 -> 0.807$\pm$0.01 | Helped |
| Use predictive clustering with outcome-conditioned centroids rather than the strongest listed baseline | AC-TPC · UKCF | ARI: 0.440$\pm$0.02 -> 0.602$\pm$0.01 | Helped |
| Use predictive clustering with outcome-conditioned centroids rather than the strongest listed baseline | AC-TPC · ADNI | Purity: 0.776$\pm$0.05 -> 0.786$\pm$0.03 | Helped |
| Use predictive clustering with outcome-conditioned centroids rather than the strongest listed baseline | AC-TPC · ADNI | AUROC: 0.756$\pm$0.04 -> 0.768$\pm$0.02 | Helped |
| Add sample-wise entropy loss to encourage dominant cluster assignments (ablation) | AC-TPC · UKCF, 3 comorbidities | Purity: 0.573$\pm$0.01 -> 0.808$\pm$0.01 | Helped |
| Add sample-wise entropy loss to encourage dominant cluster assignments (ablation) | AC-TPC · UKCF, 3 comorbidities | AUROC: 0.500$\pm$0.00 -> 0.852$\pm$0.00 | Helped |
| Add embedding separation loss to prevent centroid collapse (ablation) | AC-TPC · UKCF, 3 comorbidities | Purity: 0.795$\pm$0.01 -> 0.808$\pm$0.01 | Helped |
| Add embedding separation loss to encourage distinct label distributions (ablation) | AC-TPC · UKCF, 3 comorbidities | Activated clusters: 8.4 -> 8 | Hurt |
| Increase possible cluster count K while using β=1.0 | AC-TPC · UKCF | Average SI: 0.120 | Mixed |
Closest papers by meaning
- Learning of Cluster-based Feature Importance for Electronic Health Record Time-series (ICML 2022 · similarity 0.92)
- Clustering Interval-Censored Time-Series for Disease Phenotyping (AAAI 2022 · similarity 0.88)
- Learning Phenotypes and Dynamic Patient Representations via RNN Regularized Collective Non-Negative Tensor Factorization (AAAI 2019 · similarity 0.86)
- Deep Mixed Effect Model Using Gaussian Processes: A Personalized and Reliable Prediction for Healthcare (AAAI 2020 · similarity 0.86)
- Risk Horizons: Structured Hypothesis Spaces for Longitudinal Clinical Prediction (NeurIPS 2026 · similarity 0.84)
- Integrating Sequence and Image Modeling in Irregular Medical Time Series Through Self-Supervised Learning (AAAI 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.