Unlock: Gaussian Process Regression

Inference with Gaussian processes: the prior-to-posterior update in closed form, the role of kernel choice, marginal likelihood for hyperparameter selection, sparse approximations for scalability, and the connection to Bayesian optimization.