Regularization
A set of mathematical techniques used to artificially penalize a model's complexity during training, forcing it to generalize rather than memorize the data.
Think of It Like This
Like a strict referee giving a sprinter a heavy backpack during practice so they are forced to build genuine muscle rather than relying on a tailwind.
Techniques like L1/L2 penalties, Dropout, and early stopping all fall under the regularization umbrella. By penalizing excessively large weights, the model is prevented from fitting perfectly to the noisy outliers in the training set. It is an absolutely mandatory component for training any deep neural network successfully.