Preface

When we started planning The Kaggle Book, we had more than 85,117 novices (as they have at least just registered) and 57,426 contributors (as they have at least filled their profile) present at the time on the Kaggle platform. We wondered how to help them break the ice with data science competitions on Kaggle. We then decided to provide them with the best available information about Kaggle and data science competitions and help them to start their journey in the best possible way, thanks to hints and suggestions by over 30 Kaggle Masters and Grandmasters.

Only when we had completed our work, we realized that there was little space left in the book for anything else and that, regrettably, we had to exclude some practical demonstrations and examples. However, practice sometimes is as important as theory (we know about it very well since we are applied data scientists!), and theory cannot be considered complete without any practice. Finally, The Kaggle Workbook is here to supplement The Kaggle Book by providing you with guided exercises to put some of the ideas found in The Kaggle Book into practice.

Strictly speaking, in this workbook, you will find:

• The exploration of an emblematic selection of competitions (tabular, forecasting, computer vision, and natural language processing) where we demonstrate how a simple and effective solution can be derived for each of them.
• Reference to concepts and ideas to be found in the original Kaggle Book.
• Some challenges for the readers, as we pose some questions (and exercises) help hone your skills on the same proposed competitions or in analogous ones.

First, by reading The Kaggle Book and then practicing on The Kaggle Workbook, you’ll have all the skills, both theory-based and hands-on, necessary to compete on Kaggle for glory, fun, or learning, and gathering interesting, applied projects to present in a job interview, or for your own portfolio!

Let’s not wait; let’s start practicing now!