XGBoost (Gradient Boosting) Visualizer

Demystify sequential ensemble learning by watching weak decision trees fit to residuals step-by-step.

True Target (y)
Global Ensemble Prediction (y_hat)
Current Residuals (y - y_hat)
💡 Quick Classroom Guide: Watch how the red residual dots flatten out toward zero as you add more trees. That represents the ensemble slowly master-fitting the target curve!
Confusion Matrix (Discrete Threshold Partition Mapping)
Predicted High (+)
Predicted Low (-)
Total
Actual High
0 0.0% True High (TP)
0 0.0% False Low (FN)
0
Actual Low
0 0.0% False High (FP)
0 0.0% True Low (TN)
0
Total
0
0
0
Learning Rate (η)0.30
Tree Splits (Max Depth)Stumps (1 Split)
(Simplified to stumps to clearly visualize thresholds)
Model Initialization State
Trees Trained: 0
Current Ensemble Mean Squared Error: 0.00
Accuracy
0.0%
Precision
0.0%
Sensitivity (Recall)
0.0%
Specificity
0.0%
Model initialized with base constant prediction (mean value of y).