Why scikit-learn isn’t the Best for Visualizing Decision Trees: Meet dtreeviz

Author(s): Souradip Pal

Originally published on Towards AI.

Fall in Love with Decision Trees with dtreeviz’s Visualization

This member-only story is on us. Upgrade to access all of Medium.

Decision Trees, also known as CART (Classification and Regression Trees), are undoubtedly one of the most intuitive algorithms in the machine learning space, thanks to their simplicity. Unlike neural networks or SVMs, where you have to invest considerable time to understand the underlying processes, decision trees are essentially a series of “if-else” statements stacked together to guide you toward a possible outcome. Sure, there’s some math involved in determining those conditions, but it’s not too overwhelming.

Yes, decision trees are straightforward, but there’s a catch. The problem doesn’t lie with the algorithm itself, but with the tools often used to visualize it — specifically, scikit-learn. The visualizations produced by scikit-learn can turn you off from decision trees altogether. Why am I saying this? Well, let’s dive into an example so you can see for yourself.

To showcase the limitations, we’ll use the famous Penguin dataset, which is readily available in seaborn.

First things first, we need to import the necessary libraries to get everything rolling. Let’s get the formalities out of the way:

Here, you can see that we’ve imported all the required libraries. Now, let’s load the dataset, perform label encoding, and… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI