Cross-Validation
Machine Learning · Intermediate
Definition
K-fold cross-validation splits data into k parts, trains on k−1 and tests on the remaining one, k times.
Gives a reliable estimate of how a model performs on unseen data.
Diagram
Flowchart
Example: Average fold scores
scores = [0.82, 0.85, 0.80, 0.88, 0.85]
print(round(sum(scores) / len(scores), 3))Output: 0.84
Complexity
k times the training cost
Key points
- Stratified k-fold keeps class ratios
- Time series needs forward-chaining splits
Interview questions
Why not one train/test split?
A single split can be lucky or unlucky; k folds reduce variance.