Check your understanding
Quiz
You have learned to estimate a model's real accuracy with cross-validation, resample safely when data is grouped or time-ordered, tune without fooling yourself using nested CV, search hyperparameters efficiently, pick scoring rules that match the decision, compare models with uncertainty, and translate a metric gain into money. This quiz checks what stuck. The last two steps are live R you can run.
Question 1 of 6
Why not a single split
Compared with one train/test split, k-fold cross-validation gives:
Correct: averaging over folds reduces the luck of one particular split.
CV measures performance; it does not prevent overfitting.