ENCO
Official repository of the paper "Efficient Neural Causal Discovery without Acyclicity Constraints"
view repo
Learning the structure of a causal graphical model using both observational and interventional data is a fundamental problem in many scientific fields. A promising direction is continuous optimization for score-based methods, which efficiently learn the causal graph in a data-driven manner. However, to date, those methods require constrained optimization to enforce acyclicity or lack convergence guarantees. In this paper, we present ENCO, an efficient structure learning method for directed, acyclic causal graphs leveraging observational and interventional data. ENCO formulates the graph search as an optimization of independent edge likelihoods, with the edge orientation being modeled as a separate parameter. Consequently, we can provide convergence guarantees of ENCO under mild conditions without constraining the score function with respect to acyclicity. In experiments, we show that ENCO can efficiently recover graphs with hundreds of nodes, an order of magnitude larger than what was previously possible, while handling deterministic variables and latent confounders.
READ FULL TEXT
We consider the problem of learning causal directed acyclic graphs from ...
read it
Knowing the causal structure of a system is of fundamental interest in m...
read it
Structure learning of directed acyclic graphs (DAGs) is a fundamental pr...
read it
Recently directed acyclic graph (DAG) structure learning is formulated a...
read it
Causal structure learning has been a challenging task in the past decade...
read it
Causal discovery from observational data is an important but challenging...
read it
Discovering causal structure among a set of variables is a fundamental
p...
read it
Comments
There are no comments yet.