Epoch

An epoch is the term used for one full cycle through the entire training dataset during the training of a machine learning model.

How it works

During an epoch, the model's weights are updated to minimize the error in prediction. The number of epochs is a hyperparameter that defines the total number of times the learning algorithm will work through the entire training dataset.

The number of epochs is essentially the number of times the process of forward and backward propagation is repeated.

Here, too many epochs can lead to overfitting of the training dataset, whereas too few epochs may result in underfitting. Hence, the selection of the number of epochs is usually done through trial and error.

Keep in mind that one epoch is too big to feed to the computer at once so we divide it in several smaller batches. The number of epochs means how many times the learning algorithm will pass through the entire training dataset.