37. Regularization

Techniques used to prevent overfitting in AI models by adding a penalty to the loss function, encouraging simpler models.