用先验拟合网络提升因果结构发现的可靠性
Amortized Causal Discovery with Prior-Fitted Networks
- 用先验拟合网络替代传统似然估计，实现可泛化的因果推断
- 在合成与真实数据上，结构恢复准确率显著优于基线方法
- 特别适合需要高精度因果推理的科研与工业场景
近年来，可微分惩罚似然方法因其能通过最大化数据似然来优化因果结构而受到关注。然而，近期研究显示，即使在较大样本量下，似然估计误差也会导致无法正确发现因果结构。本文提出一种新的近似因果发现方法，利用先验拟合网络（PFNs）实现数据相关的似然估计，从而获得更可靠的结构学习评分。在合成数据、模拟数据和真实数据上的实验表明，该方法在结构恢复方面相比标准基线有显著提升。此外，我们直接证明了PFNs在似然估计准确性上优于传统神经网络方法。
原文摘要 · Abstract (English)
In recent years, differentiable penalized likelihood methods have gained popularity, optimizing the causal structure by maximizing its likelihood with respect to the data. However, recent research has shown that errors in likelihood estimation, even on relatively large sample sizes, disallow the discovery of proper structures. We propose a new approach to amortized causal discovery that addresses the limitations of likelihood estimator accuracy. Our method leverages Prior-Fitted Networks (PFNs) to amortize data-dependent likelihood estimation, yielding more reliable scores for structure learning. Experiments on synthetic, simulated, and real-world datasets show significant gains in structure recovery compared to standard baselines. Furthermore, we demonstrate directly that PFNs provide more accurate likelihood estimates than conventional neural network-based approaches.
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。