Cross-Validation and Epoch

vision
1

I am trying to understand the relationship between cross-validation and number of epoch.

Let’s say I am using 10 fold cross-validation, and now I set the number of epoch = 100.

Does that mean my model will run through the whole data by 1000 times? Then, the actual number epoch became 1000?

2

In the basic approach, called k-fold CV, the training set is split into k smaller sets (other approaches are described below, but generally follow the same principles). The following procedure is followed for each of the k “folds”:

  • A model is trained using k−1 of the folds as training data;
  • the resulting model is validated on the remaining part of the data (i.e., it is used as a test set to compute a performance measure such as accuracy).

So my answer is that the whole model will see your 10 - 1 th fold data which gets shuffled 100 times while training
and the last fold for validation (100 times) in hundred epochs.

3

i am sure one dataset will be executed 10 times in one epoch coz of 10 fold.

4

then it contradicts the actual meaning of an epoch.
An epoch in machine learning means one complete pass of the training dataset through the algorithm.

5

True, that’s why I am feeling confused. :slight_smile: