Bayesian Networks and Structural Causal Models Relationship Explored

A recent study published on arXiv (2603.27406) explores the connection between Bayesian networks and probabilistic structural causal models. It assesses if a Bayesian network, whether constructed from expert insights or generated through data, can be aligned with a probabilistic structural causal model, along with the implications for network configuration and probability distribution. The authors reveal that linear programming and linear algebra are essential techniques for this conversion, and they investigate the conditions for the existence and uniqueness of solutions, which depend on the dimensions of the probabilistic structural model.