Approximate Bayesian computation (ABC) enables more realistic modelling on high-dimensional, structured data sets by comparing a set of lower-dimensional summary statistics which are evaluated on forward-simulated data under the model and the observed data.
