📊 Classic MLID: cross-validation

K-Fold Cross-Validation

Q: Write the K-Fold error estimate $\hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k$ and explain why K=5 or 10 beats LOOCV (bias-variance angle)?

【K-Fold Cross-Validation 核心考点 1】Write the K-Fold error estimate $\hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k$ and explain why K=5 or 10 beats LOOCV (bias-variance angle)?。技术解析：Definition: K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds \hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k where E_k is the validation error of fold k. Key variants: St... 完整推导与工程落地见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/ML/card/cross-validation

Q: Stratified vs plain K-Fold? Why does ignoring stratification hurt classification? When must you use Group K-Fold?

【K-Fold Cross-Validation 核心考点 2】Stratified vs plain K-Fold? Why does ignoring stratification hurt classification? When must you use Group K-Fold?。技术解析：Definition: K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds \hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k where E_k is the validation error of fold k. Key variants: St... 完整推导与工程落地见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/ML/card/cross-validation

Q: Why does larger K reduce bias but increase variance? How does training-set overlap between folds affect variance?

【K-Fold Cross-Validation 核心考点 3】Why does larger K reduce bias but increase variance? How does training-set overlap between folds affect variance?。技术解析：Definition: K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds \hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k where E_k is the validation error of fold k. Key variants: St... 完整推导与工程落地见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/ML/card/cross-validation

Q: CV and leakage: why must scaling, imputation and feature selection be done inside each fold?

【K-Fold Cross-Validation 核心考点 4】CV and leakage: why must scaling, imputation and feature selection be done inside each fold?。技术解析：Definition: K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds \hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k where E_k is the validation error of fold k. Key variants: St... 完整推导与工程落地见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/ML/card/cross-validation

Q: What does a training error far below the CV error tell you? How does CV error relate to test error?

【K-Fold Cross-Validation 核心考点 5】What does a training error far below the CV error tell you? How does CV error relate to test error?。技术解析：Definition: K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds \hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k where E_k is the validation error of fold k. Key variants: St... 完整推导与工程落地见：https://talentme.airsota.com/resources/tech/roadmap/mindmap/ML/card/cross-validation

K-Fold 交叉验证

🎯Core Definition

K-Fold cross-validation splits the dataset into K equal folds, trains on K-1 and validates on the remaining one in rotation; the error estimate is the mean over folds

\hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k

where

E_k

is the validation error of fold k. Key variants: Stratified K-Fold (each fold keeps the global class proportion, mandatory for classification), Group K-Fold (samples of the same group — same user or experiment — never span folds, preventing group-correlation leakage), and leave-one-out (LOOCV, K=N). Bias-variance tradeoff: larger K means more training data and lower bias, but more overlap between training sets and higher variance; K=5 or 10 is the usual compromise.

💡Use Cases

estimating generalization with limited data; comparing configurations for hyperparameter tuning, model selection and feature selection (pick the best CV score); interviews ask about choosing K, stratified vs group, and how comparing CV error with training error reveals overfitting.

⚡Key Problems Solved

a single holdout split depends on luck and has high variance (the same model can differ by several points across splits); K-Fold validates every sample exactly once, giving a more stable and trustworthy estimate; compared with LOOCV (

O(N)

trainings, expensive and high variance), K-Fold trades bias against variance and is the default standard for ML evaluation and selection.

🎯5 High-Frequency Exam Points

Write the K-Fold error estimate

\hat{E} = \frac{1}{K}\sum_{k=1}^{K} E_k

and explain why K=5 or 10 beats LOOCV (bias-variance angle)?

Stratified vs plain K-Fold? Why does ignoring stratification hurt classification? When must you use Group K-Fold?

Why does larger K reduce bias but increase variance? How does training-set overlap between folds affect variance?

CV and leakage: why must scaling, imputation and feature selection be done inside each fold?

What does a training error far below the CV error tell you? How does CV error relate to test error?

📖 In-depth Guide:📄 ml-math-and-eval-metrics →

Updated 2026-08-12

🎯

Test Your Knowledge: Practice Questions for "K-Fold Cross-Validation"

Single choice pitfall questions with instant feedback and mistake tracking.

🚀 Start Card Practice ➔

← Previous CardProbability Calibration Next Card →Time-Series Cross-Validation

🔗 More Classic ML Knowledge Cards

AdaBoost Derivation Bagging & Random Forest Baum-Welch (HMM EM)GBDT Negative Gradient