arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2303.04201cs.LGstat.ME
收藏

DR-VIDAL——用于真实世界数据反事实预测与治疗效应估计的双重鲁棒变分信息论深度对抗学习

DR-VIDAL -- Doubly Robust Variational Information-theoretic Deep Adversarial Learning for Counterfactual Prediction and Treatment Effect Estimation on Real World Data

University of Pittsburgh(匹兹堡大学);University of Florida(佛罗里达大学)

Shantanu Ghosh, Zheng Feng, Jiang Bian, Kevin Butler, Mattia Prosperi

AI总结:

针对真实世界观察数据的个体治疗效应估计偏差问题,提出融合VAE、Info-GAN与双重鲁棒模块的DR-VIDAL生成框架,在多数据集上性能优于现有方法。

摘要:

由于存在潜在偏差,从真实世界的观察性(非随机)数据(例如利用电子健康记录进行药物重定位)中确定干预措施对结果的因果效应颇具挑战。在估计个体治疗效应(ITE)方面,因果深度学习相比传统技术已有所改进。我们提出了双重鲁棒变分信息论深度对抗学习(DR-VIDAL),这是一种新型生成框架,结合了两个治疗与结果的联合模型,即便其中一个模型设定有误,也能确保无偏的ITE估计。DR-VIDAL整合了三个部分:(i)变分自编码器(VAE),用于根据因果假设将混杂因素分解为潜在变量;(ii)信息论生成对抗网络(Info-GAN),用于生成反事实;(iii)纳入治疗倾向评分的双重鲁棒模块,用于结果预测。在合成数据集和真实世界数据集(婴儿健康与发展项目、双胞胎出生登记库、全国支持工作项目)上,DR-VIDAL的表现优于其他非生成式和生成式方法。综上,DR-VIDAL将因果假设、VAE、Info-GAN与双重鲁棒性独特融合,构建了一个全面且性能优异的框架。代码遵循MIT许可证,可在以下网址获取:https://github.com/Shantanu48114860/DR-VIDAL-AMIA-22。

英文摘要:

Determining causal effects of interventions onto outcomes from real-world, observational (non-randomized) data, e.g., treatment repurposing using electronic health records, is challenging due to underlying bias. Causal deep learning has improved over traditional techniques for estimating individualized treatment effects (ITE). We present the Doubly Robust Variational Information-theoretic Deep Adversarial Learning (DR-VIDAL), a novel generative framework that combines two joint models of treatment and outcome, ensuring an unbiased ITE estimation even when one of the two is misspecified. DR-VIDAL integrates: (i) a variational autoencoder (VAE) to factorize confounders into latent variables according to causal assumptions; (ii) an information-theoretic generative adversarial network (Info-GAN) to generate counterfactuals; (iii) a doubly robust block incorporating treatment propensities for outcome predictions. On synthetic and real-world datasets (Infant Health and Development Program, Twin Birth Registry, and National Supported Work Program), DR-VIDAL achieves better performance than other non-generative and generative methods. In conclusion, DR-VIDAL uniquely fuses causal assumptions, VAE, Info-GAN, and doubly robustness into a comprehensive, performant framework. Code is available at: https://github.com/Shantanu48114860/DR-VIDAL-AMIA-22 under MIT license.

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。

↑