Back to GalleryAdaBoost
AdaBoost (Adaptive Boosting) trains weak learners (usually 1-level decision stumps) sequentially. After each round, misclassified samples get higher weights, forcing the next learner to focus on difficult instances. Predictions are combined as a weighted sum.
Key Parameters & Visual Influence:
• Number of Stumps: More stumps build more intricate, non-linear boundaries, but excessive counts can overfit.
• Stump Depth: Controls decision boundary complexity of individual stumps. Depth of 1 yields axis-aligned linear partitions.
• Shrinkage: Scales the step size of stumps, acting as a regularization parameter.
Dataset Studio
CSV format: Headers in first row, last column = label/target. All values must be numeric.
Example: x1, x2, label