arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.09207cs.LGcs.AIstat.ME
收藏

使用结构化状态空间模型的反事实结果预测

Counterfactual Outcome Prediction using Structured State Space Model

Purdue University(普渡大学)

Vishal Purohit

AI总结:

本文探讨了状态空间模型在纵向数据反事实结果预测中的应用,通过实验证明S4Model比TE-CDE训练更高效、预测更准确且更稳定。

摘要:

纵向数据中的反事实结果预测因其在医疗保健和社会科学中的潜在应用而近期受到关注。本文探讨了状态空间模型(一种流行的序列模型)在此任务中的应用。具体而言,我们比较了两个模型的性能:Treatment Effect Neural Controlled Differential Equation (TE-CDE) 和结构化状态空间模型 (S4Model)。虽然 TE-CDE 使用受控微分方程来解决时间依赖性混杂问题,但它存在优化问题和训练缓慢的缺陷。相比之下,S4Model 在建模长程依赖性方面更高效且更易于训练。我们在模拟的肺肿瘤生长数据集上评估了这些模型,发现 S4Model 优于 TE-CDE,每个 epoch 的训练时间减少了 1.63 倍,归一化均方误差改善了 10 倍。此外,S4Model 在训练期间更稳定,对权重初始化的敏感性低于 TE-CDE。我们的结果表明,状态空间模型可能是纵向数据中反事实结果预测的一种有前景的方法,S4Model 为 TE-CDE 提供了一种更高效、更有效的替代方案。

英文摘要:

Counterfactual outcome prediction in longitudinal data has recently gained attention due to its potential applications in healthcare and social sciences. In this paper, we explore the use of the state space model, a popular sequence model, for this task. Specifically, we compare the performance of two models: Treatment Effect Neural Controlled Differential Equation (TE-CDE) and structured state space model (S4Model). While TE-CDE uses controlled differential equations to address time-dependent confounding, it suffers from optimization issues and slow training. In contrast, S4Model is more efficient at modeling long-range dependencies and easier to train. We evaluate the models on a simulated lung tumor growth dataset and find that S4Model outperforms TE-CDE with 1.63x reduction in per epoch training time and 10x better normalized mean squared error. Additionally, S4Model is more stable during training and less sensitive to weight initialization than TE-CDE. Our results suggest that the state space model may be a promising approach for counterfactual outcome prediction in longitudinal data, with S4Model offering a more efficient and effective alternative to TE-CDE.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑