Hongyi Li, Han Lin, Jun Xu
The Hinge Regression Tree (HRT) is a new method for creating oblique decision trees using a Newton method that improves split quality and convergence speed, outperforming traditional tree models.
Oblique decision trees are a powerful type of decision tree that can handle complex decision boundaries, but they are hard to optimize. The Hinge Regression Tree (HRT) offers a new approach by treating tree splits as a mathematical optimization problem. This method uses a technique similar to Newton's method to find the best splits quickly and reliably. Tests show that HRT creates simpler and more effective trees than traditional methods, making it a promising tool for data analysis.
