SCM-VAE: Learning Identifiable Causal Representations via Structural Knowledge
SCM-VAE: Learning Identifiable Causal Representations via Structural Knowledge
复制标题
DOI：
10.1109/bigdata55660.2022.10021114
复制
发表时间：
2022-12
期刊：
影响因子：
--
通讯作者：
Aneesh Komanduri;Yongkai Wu;Wen Huang;Feng Chen;Xintao Wu
中科院分区：
文献类型：
--
作者：
Aneesh Komanduri;Yongkai Wu;Wen Huang;Feng Chen;Xintao Wu
因果表示学习的目标是将低级观察映射到高级因果概念，以学习各种下游任务的可解释且鲁棒的表示。潜变量模型，如变分自动编码器（VAE）经常被用来学习解纠缠表示。然而，观察到的数据往往存在复杂的非线性因果关系，无法通过分解表示或线性相关假设来捕获。此外，独立的条件先验假设可以使潜在空间中的因果依赖关系的学习更具挑战性。我们提出了一个框架，创造SCM-VAE，它使用先验因果知识，结构因果先验，和非线性加性噪声结构因果模型（SCM）学习独立的因果机制和可识别的因果表示。我们进行理论分析，并在合成和真实世界的数据集上进行实验，以显示学习的因果表示的质量提高和干预下的鲁棒性。
The goal of causal representation learning is to map low-level observations to high-level causal concepts to learn interpretable and robust representations for various downstream tasks. Latent variable models such as the variational autoencoder (VAE) are frequently leveraged to learn disentangled representations. However, there are often complex non-linear causal relationships underlying the observed data that cannot be captured through disentangled representations or linear dependence assumptions. Further, an independent conditional prior assumption can make learning causal dependencies in the latent space more challenging. We propose a framework, coined SCM-VAE, which uses apriori causal knowledge, a structural causal prior, and a non-linear additive noise structural causal model (SCM) to learn independent causal mechanisms and identifiable causal representations. We conduct theoretical analysis and perform experiments on synthetic and real-world datasets to show the improved quality of learned causal representations and robustness under interventions.