Regularized Greedy Forest

What is Regularized Greedy Forest (RGF)?

Regularized Greedy Forest (RGF) is an ensemble learning method for classification and regression tasks. It is an extension of the gradient boosting algorithm and aims to improve the performance of decision tree-based models. RGF uses a greedy algorithm to construct a forest of decision trees, where each tree is built by minimizing the regularized loss function that includes the L2 regularization term.

Example of Regularized Greedy Forest

Here’s an example of using the RGFClassifier from the rgf_python package for a classification problem:

from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
from rgf.sklearn import RGFClassifier

# Load the Iris dataset
data = load_iris()
X, y =,

# Split the data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Create the Regularized Greedy Forest classifier
rgf = RGFClassifier(max_leaf=50, algorithm="RGF_Sib", test_interval=100, verbose=False)

# Train the classifier, y_train)

# Make predictions on the test set
y_pred = rgf.predict(X_test)

# Calculate the accuracy of the classifier
accuracy = accuracy_score(y_test, y_pred)
print("RGF Classifier accuracy:", accuracy)

This code would output something like:

RGF Classifier accuracy: 1.0

The Regularized Greedy Forest classifier achieves perfect accuracy on the Iris dataset.