Compare categorical encoding strategies for tree models

Compare the performance and structure (e.g., maximum depth) of the library Decision Tree when using One-Hot Encoding versus Label Encoding/Ordinal Encoding for categorical variables. Explain in a short paragraph which encoding is more efficient for tree-based splits on this specific dataset.

The issue names no notebook or test file, so first locate the library's Decision Tree implementation and the specific dataset used by the project. Run comparable One-Hot and Label/Ordinal Encoding experiments, record performance and maximum depth, and finish with the requested short explanation of which encoding is more efficient for this dataset.

Written by the indexing model from the issue text.