Meta Learning for Causal Direction
This addresses a fundamental issue in causal inference for fields where controlled randomized trials are inaccessible, though it appears incremental in combining existing meta-learning and causal inference techniques.
The paper tackles the problem of distinguishing cause from effect in bivariate settings with limited observational data, introducing a meta-learning approach that maintains high accuracy across varying dataset sizes.
The inaccessibility of controlled randomized trials due to inherent constraints in many fields of science has been a fundamental issue in causal inference. In this paper, we focus on distinguishing the cause from effect in the bivariate setting under limited observational data. Based on recent developments in meta learning as well as in causal inference, we introduce a novel generative model that allows distinguishing cause and effect in the small data setting. Using a learnt task variable that contains distributional information of each dataset, we propose an end-to-end algorithm that makes use of similar training datasets at test time. We demonstrate our method on various synthetic as well as real-world data and show that it is able to maintain high accuracy in detecting directions across varying dataset sizes.