Yellowbrick – A set of Visualization Tools to Accelerate your Model Selection Process

Overview

• Yellowbrick is a suite of visualization and diagnostic tools that will enable quicker model selection
• It’s a Python package that combines scikit-learn and matplotlib
• Some of the more popular visualization tools include model selection, feature visualization, classification and regression visualization

Introduction

So you’ve built a bunch of models for your project (or hackathon). A few of them are even giving you a really good AUC-ROC score. But how do you go about choosing one? Sure you can get into the nitty gritty of it but wouldn’t it help if you could just visualize them and pick the best one from there?

Yellowbrick is a suite of visualization tools that extend the scikit-learn API to enable acceleration and ease of model selection for data scientists. In simple words, this Python package combines the power of scikit-learn with the capabilities of matplotlib to generate intuitive visualizations of your models.

This set of tools are called Visualizers, which are basically objects that learn from the data and then create visualizations so you can gain a deeper insight into the model selection process. Some of the most popular visualizers are listed below:

• Model Selection Visualization
• Regression Visualization
• Classification Visualization
• Feature Visualization
• Clustering Visualization
• Text Visualization

Yellowbrick is compatible with Python 2.7 as well but the developers recommend using Python 3.5 to utilise this package to its maximum potential. Of course you will also need to install scikit-learn and matplotlib as well. To install Yellowbrick, use the pip method:

$ pip install yellowbrick

To see example of Yellowbrick in action and to replicate what the developers have demonstrated, head over to the GitHub page here.

Our take on this

I love this package! Adding visualization to any part of the data science process always helps in my opinion. We have previously covered the ANN visualiser which illustrates the artificial neural network you build and we have also seen the deep learning process being visualized.

Yellowbrick will really help you, as a data scientist, dig deeper into your model understanding and also save you time by portraying the process in the form of a neat diagram. I urge you to try out this library and let us know your experience using it.

Subscribe to AVBytes here to get regular data science, machine learning and AI updates in your inbox!

Pranav Dar

Senior Editor at Analytics Vidhya.Data visualization practitioner who loves reading and delving deeper into the data science and machine learning arts. Always looking for new ways to improve processes using ML and AI.