Machine Learning With Go

Overview of decision trees and random forests

Again, consider our classes A and B. In this case, suppose that we have one feature, x₁, that ranges from 0.0 to 1.0, and we have another feature, x₂ that is categorical and can take on one of two values, a₁ and a₂ (this could be something like male/female or red/blue). A decision tree to classify a new data point might look something like the following:

There are a variety of ways to choose how decision trees are constructed, split, and so on. One of the most common ways to determine how decision trees are constructed is using a quantity called entropy. This entropy-based approach is discussed ...