Simulation-based inference approach emulates traditional Bayesian inference approach. When assessing the parameters of a model, one first defines prior distributions and then defines the likelihood of a given observation, often using a forward-modeling approach. This likelihood is further sampled to obtain the posterior distribution of the parameters. The simulation-based approach does not require explicit computation of the likelihood, and instead it will learn an approximation of the desired distribution (i.e., the likelihood or directly the posterior distribution) by training a neural network with a sample of simulated observations.