Variational inference is a family of optimisation-based methods for approximating complex posterior distributions in Bayesian models. By transforming inference into an optimisation problem, these ...