Epoch

In the context of training machine learning models, particularly neural networks, an "epoch" represents one complete pass of the entire training dataset through the learning algorithm. During a single epoch, every example in the training set is presented to the model at least once, allowing the model to make predictions, calculate the error (loss), and update its internal parameters (weights and biases) based on that error. Because individual data points are often processed in smaller subsets called "batches" for computational efficiency, a full epoch consists of multiple iterations, where each iteration processes one batch.