I want to compare my model on a train_test_split data vs. KFold data. My understanding is that if I have 5 folds, I will split the data 5 ways in a way where a different 20%
