Generally want the most probable hypothesis given the training data
Maximum a posteriori hypothesis 8#8:
If assume 13#13 then can further simplify, and choose the Maximum likelihood (ML) hypothesis
