When we do k-fold cross validation, should we just use the classifier that has the highest test accuracy? What is generally the best approach in getting a classifier from cross validation?
tuomastik
- 1.2k
- 10
- 22