Causal Modeling with Stationary Diffusions

arXiv:2310.17405v2 Announce Type: replace
Abstract: We develop a novel approach towards causal inference. Rather than structural equations over a causal graph, we learn stochastic differential equations (SDEs) whose stationary densities model a system's behavior under interventions. These stationary diffusion models do not require the formalism of causal graphs, let alone the common assumption of acyclicity. We show that in several cases, they generalize to unseen interventions on their variables, often better than classical approaches. Our inference method is based on …

abstract arxiv behavior causal causal inference cs.lg differential diffusion diffusion models graph graphs inference learn modeling novel stochastic type