用专家模型引导扩散模型，提升复杂系统反事实预测的可靠性。
Counterfactual Probabilistic Diffusion with Expert Models
- 通过提取专家模型的高层信号作为生成先验，融合机理与数据驱动方法
- 在新冠模拟、药理动态等场景中，点预测与分布精度均优于主流基线
- 适合需要可解释因果推断的医疗健康领域研究者
在公共卫生与医学等领域的科学建模和决策中，预测复杂动力系统的反事实分布至关重要。然而，现有方法多依赖点估计或纯数据驱动模型，在数据稀缺时表现不佳。本文提出基于时间序列扩散的框架 ODE-Diff，通过提取不完美专家模型的高层信号作为结构化先验，实现机理模型与数据驱动方法的融合，提升因果推断的可靠性与可解释性。我们在半合成新冠模拟、合成药理动力学及真实案例研究中评估该方法，结果表明其在点预测与分布准确性上持续优于强基线模型。
原文摘要 · Abstract (English)
Predicting counterfactual distributions in complex dynamical systems is essential for scientific modeling and decision-making in domains such as public health and medicine. However, existing methods often rely on point estimates or purely data-driven models, which tend to falter under data scarcity. We propose a time series diffusion-based framework that incorporates guidance from imperfect expert models by extracting high-level signals to serve as structured priors for generative modeling. Our method, ODE-Diff, bridges mechanistic and data-driven approaches, enabling more reliable and interpretable causal inference. We evaluate ODE-Diff across semi-synthetic COVID-19 simulations, synthetic pharmacological dynamics, and real-world case studies, demonstrating that it consistently outperforms strong baselines in both point prediction and distributional accuracy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。