Real-world uses

Created for this library

1. 1.

   A risk modeling team frames its training as empirical risk minimization on a chosen surrogate loss to get a tractable optimization problem.

2. 2.

   An ML platform team teaches its new engineers ERM as the conceptual basis for supervised training to anchor advanced topics later.

3. 3.

   A research team contrasts ERM with distributionally robust alternatives when training a model expected to face shifts in the input distribution.