Machine Learning Fundamentals · free preview
Data Splits and K-Fold Cross-Validation
Model selection needs a quantity the training set cannot provide, so the data is split three ways, and K-fold cross-validation recovers a usable validation estimate when data is scarce.
1. The problem selection creates
The problem selection creates
2. The three-way split
The three-way split
3. K-fold cross-validation
K-fold cross-validation
4. What the validation set is for
In a three-way split, the quantity used to compare candidate models and tune hyperparameters is the ____ error.
5. What K-fold evaluates on
In -fold cross-validation with , each round trains on ____ folds and validates on ____ fold.
6. Why the test set stays untouched
A practitioner compares 300 candidate hyperparameter settings by evaluating each on the test set and picking the best. Which statement describes the consequence?
Practice this interactively
Free account · instant grading · spaced review that schedules itself.
Start this course for free