Tree-based Ensemble Models

---

Topics and Objectives

• Random forests and kernel view
• Generalized random forests via influence function split
• Uncertainty quantification with U-statistics
• Variable importance and variations
• Adaboost and gradient boosting
• Boosting training error bound

---

Week 5 (Sep 23 and Sep 25)

• Sep 23: [Bias-variance of Kernel Smoother]
• Sep 23: [Tree model algorithm review]
• Sep 25: [Random Forest Induced Kernel]
• Sep 25: [Random Forests and Adaptiveness]

---

Homework 3

• [Homework 3] [Source Code]
• Due Oct 15, 11:59PM CT

---

Week 6 (Sep 23 and Sep 25)

• Sep 30: [AdaBoost and Training Error Bound]
• Oct 2: [Gradient Boosting Machine]

---

Week 7 (Sep 30 and Oct 2)

• Oct 7: [Generalized Random Forests and Influence Function]
• Oct 9: [Uncertainty Quantification with Resampling Approaches]

---