Research map / Probabilistic and scientific modeling / Dynamical systems
Scalable Variational Inference for Dynamical Systems
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 mean-field variational gradient matching instead of spline gradient matching | Lotka-Volterra · Lotka-Volterra | parameter estimation accuracy | Helped |
| Use mean-field variational gradient matching instead of adaptive gradient matching | Lotka-Volterra · Lotka-Volterra | runtime: 4.7sec; roughly 100 times the runtime | Helped |
| Use mean-field variational gradient matching instead of adaptive gradient matching | Lotka-Volterra · Lotka-Volterra | parameter estimation accuracy | Helped |
| Infer states and parameters jointly with mean-field variational gradient matching | Protein signalling transduction pathway · Protein signalling transduction pathway, state S unobserved | parameter estimation accuracy | Helped |
| Increase observation noise from variance σ² = 0.01 to σ² = 0.1 | Protein signalling transduction pathway · Protein signalling transduction pathway | parameter estimation accuracy: σ² = 0.01; σ² = 0.1 | Hurt |
| Leave state S unobserved (ablation) | Mean-field GM · Protein signalling transduction pathway, σ² = 0.1 | parameter estimation accuracy: σ² = 0.1 | No clear effect |
| Apply the state-variable transformation that makes the equations monomial in states (ablation) | Protein signalling transduction pathway · Protein signalling transduction pathway | parameter estimation accuracy | Mixed |
| Use mean-field GM to initialize numerical integration | Bayesian numerical integration · Protein signalling transduction pathway, state S unobserved | parameter estimation error | Helped |
| Increase the system size to 1000 states | Mean-field GM · Lorenz 96 | runtime: 1000 states; less than 400 seconds | Helped |
Closest papers by meaning
- Fast Parameter Inference in Nonlinear Dynamical Systems using Iterative Gradient Matching (ICML 2016 · similarity 0.92)
- Efficient and Flexible Inference for Stochastic Systems (NeurIPS 2017 · similarity 0.90)
- SING: SDE Inference via Natural Gradients (NeurIPS 2025 · similarity 0.89)
- Closing the Approximation Gap in Simulation-free Latent SDEs (NeurIPS 2026 · similarity 0.89)
- Particle Semi-Implicit Variational Inference (NeurIPS 2024 · similarity 0.87)
- GAD-PVI: A General Accelerated Dynamic-Weight Particle-Based Variational Inference Framework (AAAI 2024 · similarity 0.86)
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.