Cross validation model

If I were to apply a cross validation model that classified my company’s malfunctioning equipments

Out of the following what approach would I use?

  1. Hide the training set while measuring sensitivity?

  2. Or Hide the testing set while measuring sensitivity?

  3. Or hide the training set while measuring specificity?

  4. Hide the test set while measuring specificity?

  5. Or would I hide the training set while measuring precision.

  6. Or should I hide the testing set while measuring precision?

Could you kindly help me figure which are the right options?

Which options out of these would be correct?

I do not have a clear picture of exactly what you intend to achieve.

But from what I know about cross validation, you use cross validation on a data set that is not very large. And you do not need to split the training data into train set and validation set. You test the accuracy of the model using testing data.

You should not train a model with both your training and testing data. In fact you should exclude all features that causes data leakage from your model.