Learning DAGs without imposing acyclicity

06/04/2020
by   Gherardo Varando, et al.
0

We explore if it is possible to learn a directed acyclic graph (DAG) from data without imposing explicitly the acyclicity constraint. In particular, for Gaussian distributions, we frame structural learning as a sparse matrix factorization problem and we empirically show that solving an ℓ_1-penalized optimization yields to good recovery of the true graph and, in general, to almost-DAG graphs. Moreover, this approach is computationally efficient and is not affected by the explosion of combinatorial complexity as in classical structural learning algorithms.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset