32. Which method is used for avoiding overfitting in a neural network?

Dropout refers to a type of regularization technique used for avoiding overfitting in a neural network. In this, randomly selected neurons are dropped during training.