Random Forest
An ensemble that averages many decorrelated decision trees trained on bootstrap samples.
- Bagging + random feature subsets per split decorrelate trees.
- Averaging drops variance while trees stay low-bias.
- Free generalization estimate via out-of-bag samples.
When to use: A reliable default for tabular data: little tuning, robust to outliers, free feature importance.
Example: 500-tree RF on a 100k-row churn dataset: each tree sees a different bootstrap and √p features per split; predictions are averaged probabilities.