Caesar AI Atlas

Dropout

Caesar AI Atlas Definition

Dropout is a neural network regularization technique that randomly deactivates selected units during training. By preventing units from relying too heavily on specific other units, it helps reduce overfitting and encourages more robust representations.

Also Referenced In

Related Terms