A data scientist trains a regression model to predict energy consumption for a smart building. The model achieves very low error on the training data but performs significantly worse on a held-out validation set. Which technique would most directly address this problem?

Feature engineering

Feature engineering can improve model performance, but it does not directly address overfitting caused by excessive complexity.

Cross-validation

Cross-validation is a method to estimate model performance on unseen data, but it does not inherently reduce overfitting; it helps detect it.

Hyperparameter tuning

Hyperparameter tuning can adjust model complexity, but regularization is a specific technique that directly penalizes overfitting.

What does this AI-900 question test?

Static NAT maps one inside address to one outside address.

What is the correct answer to this question?

The correct answer is: Regularization — The model is overfitting, meaning it has learned patterns specific to the training data that do not generalize. Regularization (e.g., L1, L2) directly penalizes large coefficients, reducing model complexity and mitigating overfitting. Cross-validation helps evaluate generalization but does not directly fix overfitting; feature engineering and hyperparameter tuning may help but are not as direct as regularization.

What should I do if I get this AI-900 question wrong?

Then try more questions from the same exam bank and focus on understanding why the wrong options are tempting.