xgboost regression demo

xgboostregressiondemo.py

import numpy as np

# Toy dataset
X = np.array([1, 2, 3, 4, 5])
y = np.array([1, 2, 3, 4, 5])

# Initial prediction: mean of y (Predictor Function)
initial_prediction = np.mean(y)
y_pred = np.array([initial_prediction] * len(y))

# Learning rate
lr = 0.1

# Boosting rounds
for round in range(1, 3):  # Two rounds for demonstration
    print(f"Boosting round: {round}")

    #So at each boosting round, the residuals are computed, a weak learner is fit on them, and the predictions are updated to reduce the MSE loss.
    residuals = y - y_pred

    # 2. Fit residuals to the input features (in our simple case, fit means directly assigning)
    # Normally, this step would involve training a decision tree.
    tree_output = residuals

    # 3. Update predictions (Predictor Function)
    y_pred += lr * tree_output

    # Print updated predictions
    print(f"Updated predictions: {y_pred}")

# Final predictions
print(f"Final Predictions: {y_pred}")

'''
Boosting round: 1
Updated predictions: [2.8 2.9 3.  3.1 3.2]
Boosting round: 2
Updated predictions: [2.62 2.81 3.   3.19 3.38]
Final Predictions: [2.62 2.81 3.   3.19 3.38]
'''