Cross-Validation
A model validation technique for assessing how the results of a statistical analysis will generalize to an independent dataset.
Definition
Involves partitioning data into k folds, training on k-1 folds, and testing on the held-out fold; repeating k times. This gives robust estimates of model generalization and reveals overfitting. Governance use cases include standardizing cross-validation protocols (fold size, random seeds) across teams and documenting results for compliance audits.
Real-World Example
A marketing analytics team uses 10-fold cross-validation on its customer-churn model. Each fold produces AUC scores; they report the mean and variance to stakeholders. When variance is high, they collect additional data rather than deploying a model with unstable performance.
.svg)
