What is the purpose of pruning a decision tree?

Pruning a decision tree is useful for reducing the model's overfitting. It removes parts of the tree that fit the training data too closely but generalize poorly to new data.

Advantages of pruning:

• Reduces model complexity, which increases its generalization ability.
• Lowers the risk of overfitting.
• Improves the interpretability of the model due to a smaller tree size.

Example: after building a full tree, branches that do not significantly improve validation set performance can be removed, thereby enhancing the model's stability.