/CoCo

An optimization-based algorithm to accurately estimate the causal effects and robustly predict under distribution shifts. It leverages the invariance of causality over multiple environments.

Primary LanguagePythonMIT LicenseMIT

No issues in this repository yet.