DAG Estimation at ICML 2024

We have 2 papers accepted on DAG estimation using optimal transport at ICML 2024:

In these papers we present optimal transport approaches to DAG estimation, including parameter and structure learning under missing data settings. The significance of DAG estimation in causal discovery is well known and documented but we believe our approach presents a better way of dealing with this long-standing problem. This work is lead by our amazing student Vy Vo and in colloboration with our great colleagues at Monash.