The mlr3 ecosystem is a collection of R packages for machine learning. The base package mlr3 only provides the basic building blocks for machine learning. The extensions packages extent mlr3 with functionality for specialized tasks. The packages are listed bellow with a short description. For more information on the packages, check out their reference manuals. The dot next to the package name indicates the lifecycle stage. Packages with a green dot are stable. Experimental packages are marked with an orange dot .
Core
Basic building blocks for machine learning.
Meta-package intended to simplify both installation and loading of packages from the mlr3 ecosystem.
Optimization
Hyperparameter tuning for mlr3 learners.
Collection of search spaces for hyperparameter tuning.
Successive halving and hyperband tuner for mlr3tuning.
Model-based optimization for mlr3tuning.
Flexible mixed integer evolutionary strategies.
Automated machine learning.
Feature Selection
Filter feature selection.
Wrapper feature selection.
Data
Data backend to transparently work with databases.
Learners
The mlr3 ecosystem
Essential learners for mlr3, maintained by the mlr-org team.
Extra learners for mlr3, implemented by the community.
Deep learning with Keras.
Visualization
Visualizations for tasks, predictions, resample results and benchmarks.
Pipelines
Dataflow programming toolkit.
Tasks
Spatiotemporal resampling and visualization methods.
Probabilistic predictions.
Storing and working with multi-output tasks.
Spatial data backends and prediction functions.
Utilities
Black-box optimization toolkit.
Universal parameter space description and tools.
Miscellaneous helper functions.
Performance measures for supervised learning.
Analysis and tools for benchmarking.
Machine learning fairness.
Connector between mlr3 and batchtools.