Definition
Random forests and boosting are ensemble methods that combine hundreds of decision trees into one more accurate model. A random forest builds trees independently on random subsets of data and variables and averages (votes) the result; boosting builds trees sequentially, each correcting the errors of the previous ones. Both rank among the most successful methods for tabular data.
Data mining & machine learning
Random forests are robust to overfitting and noise and give good results almost without tuning; boosting (stochastic gradient boosting) is often even more accurate but more sensitive to settings (number of trees, learning rate, depth). Both provide predictor importance, which partly replaces the lost interpretability.
They are used for churn prediction, risk scoring, predictive maintenance or defect detection. Always verify accuracy on test data or by cross-validation, and with imbalanced classes watch AUC and sensitivity rather than overall accuracy.
In Statistica
The Data Mining menu offers Random Forests and Boosted Trees (stochastic gradient boosting) for classification and regression, with automatic stopping by test error, predictor importance plots, a confusion matrix, ROC curve and model saving for deployment; in the Workspace environment they are easily compared with other models.
Related terms
- Decision treesA decision tree splits data by a sequence of simple rules ("age > 45 and income < 30 thousand") into ever more…
- Neural networksAn artificial neural network is a model built from layers of connected "neurons" that learns from data to approximate…
- Cross-validationCross-validation estimates how well a model will perform on new data.
- OverfittingOverfitting happens when a model learns not only the real relationships but also the random noise of the training data…
- ROC curve and AUCThe ROC curve shows how a classifier's sensitivity (share of correctly caught positives) and false-positive rate change…
Knowledgebase guides
FAQ
- When a random forest and when boosting?
- A random forest as a reliable first model without tuning; boosting when you want to squeeze out maximum accuracy and have time to tune parameters with cross-validation.
- How do I explain a "black box" model?
- Predictor importance plots and partial dependence plots show which variables affect the outcome and in which direction.
Try it on your own data
Statistica free for 30 days
Full version, no credit card. Or get a pricing estimate in a minute.
Data mining & machine learning
Updated: September 2026.