Why R? Webinar 009 B.Bischl+F.Pfisterer+M.Binder: Pipelines and AutoML with mlr3

Donate http://whyr.pl/donate/

• stream https://www.youtube.com/watch?v=4r8K3GO5wk4
• join Why R? Slack whyr.pl/slack/
• format: one 45 minutes long stream + 10 minutes for Q&A
• comments: ask questions on YouTube live chat
• channel: youtube.com/c/WhyRFoundation
• date: every Thursday 8:00 pm GMT+2

mlr3pipelines is an object-oriented dataflow programming toolkit for machine learning in R6. It provides an expressive and intuitive language to define ML workflows as directed acyclic graphs that represent data flows between computational units, e.g., data preprocessing and model fitting. Many complex ML concepts for which special purpose packages are usually provided can now be expressed in few lines of graph definition code, e.g. unions of feature views, bagging, stacking and hurdle models. Resulting pipelines are parameterized, so all components can be jointly tuned to obtain an optimal configuration.
Graphs can contain "branching" nodes which allows selective, conditional processing of execution paths. The tuning of such tasks makes complex model selection possible.
The modular, object-oriented concept of mlr3pipelines facilitates convenient extension with custom operations, while the compatibility with mlr3 allows convenient tuning, benchmarking, nested resampling and more.

This leads to the possibility of specifying custom, minimal automatic machine learning systems by constructing and tuning pipelines that can be adjusted for a respective task.

Project page: https://github.com/mlr-org/mlr3pipelines