Oct, 2023

通过观测变量分组使因果表示学习具备可辨识性

TL;DRCausal Representation Learning (CRL) is an ill-posed problem combining representation learning and causal discovery, and this paper presents a novel approach based on weak constraints and observational mixing, which achieves identifiability without temporal structure, intervention, or weak supervision. The paper also introduces a self-supervised estimation framework that outperforms existing baselines in CRL and demonstrates robustness against latent confounders and causal cycles.