Chapter 5
Decision Trees and Random Forests
Section 4
Using Decision Trees: Advantages and Predicted Probabilities
The logistic regression has a linear decision boundary, which will be the straight line between the lightest blue and red patches in the background. The logistic regression decision boundary goes right through the middle of the data and doesn't provide a useful classifier. This shows the power of decision trees "out of the box", without the need for engineering non-linear or interaction features. Here are the topics that we will cover now: