- Abstract: We introduce CGNN, a framework to learn functional causal models as generative neural networks. These networks are trained using backpropagation to minimize the maximum mean discrepancy to the observed data. Unlike previous approaches, CGNN leverages both conditional independences and distributional asymmetries to seamlessly discover bivariate and multivariate causal structures, with or without hidden variables. CGNN does not only estimate the causal structure, but a full and differentiable generative model of the data. Throughout an extensive variety of experiments, we illustrate the competitive esults of CGNN w.r.t state-of-the-art alternatives in observational causal discovery on both simulated and real data, in the tasks of cause-effect inference, v-structure identification, and multivariate causal discovery.
- TL;DR: Discover the structure of functional causal models with generative neural networks
- Keywords: Causal structure discovery, Generative neural networks, Cause-effect pair problem, Functional causal model, Maximum Mean Discrepancy, Structural Equation Models