Comparing nested likelihoods

Predictive evaluation asks how well a model predicts observations that were not used for fitting. We compare models using loss functions and cross-validation.

Available Wednesday, October 14, 2026.