Reinforcement Learning for Causal Discovery without Acyclicity Constraints

TL;DR

This paper introduces ALIAS, a reinforcement learning-based method for causal discovery that efficiently generates DAGs without explicit acyclicity constraints, achieving superior performance on synthetic and real data.

Contribution

ALIAS employs a novel DAG parametrization and policy gradient approach to bypass acyclicity constraints, enabling efficient and effective causal discovery.

Findings

ALIAS outperforms existing methods on synthetic datasets.

ALIAS demonstrates strong results on real-world data.

The method achieves quadratic complexity in DAG generation.

Abstract

Recently, reinforcement learning (RL) has proved a promising alternative for conventional local heuristics in score-based approaches to learning directed acyclic causal graphs (DAGs) from observational data. However, the intricate acyclicity constraint still challenges the efficient exploration of the vast space of DAGs in existing methods. In this study, we introduce ALIAS (reinforced dAg Learning wIthout Acyclicity conStraints), a novel approach to causal discovery powered by the RL machinery. Our method features an efficient policy for generating DAGs in just a single step with an optimal quadratic complexity, fueled by a novel parametrization of DAGs that directly translates a continuous space to the space of all DAGs, bypassing the need for explicitly enforcing acyclicity constraints. This approach enables us to navigate the search space more effectively by utilizing policy gradient methods and established scoring functions. In addition, we provide compelling empirical evidence for the strong performance of ALIAS in comparison with state-of-the-arts in causal discovery over increasingly difficult experiment conditions on both synthetic and real datasets.