Inference Without Exact Likelihoods

As models become increasingly realistic and data sets become more complex, likelihood functions often become difficult or impossible to evaluate. Inference for such models --- some partially and some fully intractable --- is therefore one of the major challenges of modern statistical computing. This session will showcase a variety of algorithms for this challenging class of problems, providing practical insights in the context of applications in a variety of fields, including astrophysics, environmental science, and disease modeling.