Visualising Decision Trees
Introduction Decision trees are one of the oldest and most popular forms of machine learning used for classification and regression. It’s unsurprising, then, that there’s a lot of content about them. However, most of it seems to focus on how the algorithms work, covering areas such as Gini impurity or error-minimisation. While this is useful knowledge, I’m more interested in how best to use decision trees to get the results I want - after all, my job doesn’t involve reinventing the tree, only growing them. Additionally, decision trees are some of the most easily-visualised machine learning techniques, providing high interpretability, yet often content is primarily textual, with minimal, if any, graphics.
