Hey everyone!

We are a medium-sized group of people who like devoting our free time writing open-source R packages related to machine learning.

The beginnings

It all started in 2012 when Bernd Bischl wrote the {mlr} package during his PhD.
It quickly became one of the biggest machine learning packages in R, providing a consistent syntax for handling machine learning algorithms in R.

With R being a language focusing on statistics from its origin, hundreds of algorithms exist across many packages.
Most of them were contributed by different users.
Due to the lack of API definitions, many people created slightly different interfaces for accessing the respective algorithms.

{mlr} was one of the first packages which tried to harmonize this messy landscape.
Also, it provided functionality for pre-and post-processing of data after the models have been trained.
Many people contributed to the package, especially Michel Lang, who can be seen as the co-author of the project besides Bernd.

mlr turns into mlr3

Until 2017, {mlr} has grown into a very large package (~ 40k lines of code, wrapping almost 100 R packages).
It was decided that a fresh start is needed, splitting the functionality of {mlr} into child-packages to simplify maintenance and future development.
All the knowledge that was gained through the years and new developments in the R community should be used to create an object-oriented, flexible successor of {mlr}: welcome {mlr3}!
{mlr3} was released in summer 2019 at the useR! conference in Toulouse.
At the same time, {mlr} was labelled as "retired".
The name "mlr3" originated from the fact that the now retired {mlr} package was at version 2.x at this time.

The mlr-org team

The mlr-org team consists of around 20 people, with 5-6 people forming the core team.
The core-team is responsible for the majority of contributions and discusses the future development of the package.

Sponsor contributions

With sponsor money, we will be able to cover our domain costs and organize workshops to further push the mlr3 ecosystem (we will report on workshops which were made possible due to sponsor contributions!).
If you want to become a member of our non-profit mlr-org organization, please contact [email protected].

NB: If you live in Germany, Austria or Switzerland, you can also support us without spending your hard-earned money. We are listed as a non-profit organization on https://smile.amazon.de and receive 0.5% of your purchases. Visit https://smile.amazon.de/ch/143-219-01426 to set things up.