Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Individualized Dosing Dynamics via Neural Eigen Decomposition
Stav Belogolovsky, Ido Greenberg, Danny Eytan, Shie Mannor
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 hypernetwork over patient context features (ablation) | NESDE · UH dosing | MSE: 529.7 -> 411.2 | Improved |
| add hypernetwork over patient context features (ablation) | NESDE · UH dosing | NLL: 5.42 -> 4.43 | Improved |
| add hypernetwork over patient context features (ablation) | NESDE · Vancomycin dosing | MSE: 87.32 -> 70.71 | Improved |
| add hypernetwork over patient context features (ablation) | NESDE · Vancomycin dosing | NLL: 3.73 -> 3.69 | Improved |
| decouple control from observations (NESDE) vs concatenating control as model output (GRU-ODE-Bayes) | NESDE · Synthetic complex dynamics, OOD control | MSE: 0.361 -> 0.178 | Improved |
| change control policy from negative to positive correlation with observations (OOD test) | NESDE · Synthetic complex dynamics | MSE: 0.176 -> 0.178 | No clear effect |
| change control policy to OOD test | LSTM · Synthetic real dynamics | MSE: 0.381 -> 2.354 | Hurt |
| change control policy to OOD test | CRU · Synthetic complex dynamics | MSE: 0.233 -> 0.584 | Hurt |
| reduce training set size (down to 100 trajectories) | NESDE · Synthetic benchmarks | test MSE | No clear effect |
| reduce training trajectories in OU benchmark (no control) | GRU-ODE-Bayes · OU process | loss | Hurt |
| reduce observations per training trajectory (10 down to 4) | NESDE · Synthetic complex and real dynamics | test MSE | No clear effect |
| increase LSTM prediction resolution (1, 8, 50 predictions per observation) | LSTM · Synthetic regular observations | MSE | Hurt |
Closest papers by meaning
- Bayesian Neural Controlled Differential Equations for Treatment Effect Estimation (ICLR 2024 · similarity 0.87)
- Conservative Continuous-Time Treatment Optimization (NeurIPS 2026 · similarity 0.84)
- Contrastive Balancing Representation Learning for Heterogeneous Dose-Response Curves Estimation (AAAI 2024 · similarity 0.84)
- ODE Discovery for Longitudinal Heterogeneous Treatment Effects Inference (ICLR 2024 · similarity 0.84)
- Continuous Treatment Effect Estimation Using Gradient Interpolation and Kernel Smoothing (AAAI 2024 · similarity 0.84)
- Observable Neural ODEs for Identifiable Causal Forecasting in Continuous Time (NeurIPS 2026 · 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.
