How does feature selection impact model generalization?

Feature selection reduces overfitting by removing irrelevant or noisy features.

By reducing dimensionality, feature selection decreases model variance and improves generalization. It also reduces computational cost and improves interpretability. However, removing informative features can increase bias.