Epoch

In machine learning, an epoch refers to one complete pass through the entire training dataset by the learning algorithm. During an epoch, the model's weights are updated in an attempt to minimize errors and improve accuracy.