Flexible and Robust Machine Learning Using mlr3 in R

Getting Started

Editors

Bernd Bischl, Raphael Sonabend, Lars Kotthoff, Michel Lang

Contributors

• Marc Becker
• Przemysław Biecek
• Martin Binder
• Bernd Bischl
• Lukas Burk
• Giuseppe Casalicchio
• Susanne Dandl
• Sebastian Fischer
• Natalie Foss

• Lars Kotthoff
• Michel Lang
• Florian Pfisterer
• Damir Pulatov
• Lennart Schneider
• Patrick Schratz
• Raphael Sonabend
• Marvin N. Wright

Welcome to the Machine Learning in R universe. This is the online version of the upcoming book Flexible and Robust Machine Learning Using mlr3 in R. This book will teach you about the mlr3 universe of packages, from machine learning methodology to implementations of complex algorithmic pipelines. We will cover how to use the mlr3 family of packages for data processing, fitting and training of machine learning models, tuning and hyperparameter optimization, feature selection, pipelines, data preprocessing, and model interpretability. In addition we will look at how our interface works beyond classification and regression settings to other fields including survival analysis, clustering, and more. Finally, for readers interested in technical details, we will look at implementation and decision decisions, as well as error handling and parallelization.

We hope you enjoy reading our book and always welcome comments and feedback. If you notice any mistakes in the book we would appreciate if you could open an issue in the mlr3book issue tracker. All content in this book is licensed under CC BY-NC-SA 4.0.

Community links

The mlr community is open to all and we welcome everybody, from those completely new to machine learning and R to advanced coders and professional data scientists.

The mlr3 GitHub is a good starting point for links to cheatsheets, documentation, videos, slides, overview tables, and pointers to other packages. If you want to chat to us, you can reach us on our Mattermost. For case studies and how-to guides, check out the mlr3gallery.

We appreciate all contributions, whether they are bug reports, feature requests, or pull requests that fix bugs or extend functionality. Each of our GitHub repositories includes issues and pull request templates to ensure we can help you as much as possible to get started. Please make sure you read our code of conduct and contribution guidelines before opening your first issue or pull request.

With so many packages in our universe it may be hard to keep track of where to open issues, as a general rule:

1. If you have a question about using any part of the mlr3 ecosystem, ask on StackOverflow and use the tag #mlr3 – one of our team will answer you there. Be sure to include a reproducible example (reprex) and if we think you found a bug then we will either refer you to the relevant GitHub repository or we will open an issue for you.
2. Issues or pull requests about core functionality (train, predict, etc.) should be opened in the mlr3 GitHub repository.
3. Issues or pull requests about learners should be opened in the mlr3extralearners GitHub repository.
4. Issues or pull requests about measures should be opened in the mlr3measures GitHub repository.
5. Issues or pull requests about specialized functionality (e.g., pipelines and tuning) should be opened in the GitHub repository of the respective package.

Do not worry about opening an issue in the wrong place, we will transfer it to the right one.