Decision Tree Splits - Interactive Visualization

A decision tree partitions the feature space with axis-aligned cuts, greedily choosing the split that maximally reduces impurity (measured by entropy or Gini) at each step. Each internal node asks a yes/no question about one feature, and leaf nodes hold the class prediction or regression value. Controlling tree depth is the primary lever against overfitting.

• See each split drawn on the 2D feature space and the corresponding node added to the tree diagram
• Compare entropy vs Gini impurity as splitting criteria - watch how they affect tree structure
• Grow the tree to full depth and see it memorize training data, then prune it back to reduce overfitting
• Understand information gain: the tree always picks the question that reduces uncertainty the most
• Learn why shallow trees are interpretable and why deep trees overfit - the fundamental bias-variance tradeoff in tree models

Part of the EngineersOfAI Interactive 3D - free interactive visualizations covering every major concept in machine learning and AI engineering. Hover any element for a plain-English explanation. No code required.