mlr-org
  • Overview
    • Ecosystem
    • Dependencies

    • Tasks
    • Learners
    • Resamplings
    • Measures

    • Pipeline Operators
    • Graphs
    • Torch Pipeline Operators

    • Tuners
    • Tuning Spaces
    • Terminators

    • Feature Selection Filter
    • Feature Selection Wrapper
  • Resources
  • Book
  • Gallery
  • Benchmarks
    • mlr3
    • mlr3fselect
    • mlr3tuning async
  • Team
  • More
    • Support
    • Contributing
    • Frequently Asked Questions
    • Blogroll
    • Gallery RSS

The website features runtime and memory benchmarks of the mlr3 base package now.

Gallery

In the gallery, you find case studies and demos. The posts are mostly about specific features and cover advanced topics. If you are completely new to mlr3 or machine learning, you should start with the book. Pick a post from one of the four categories or browse all posts sorted by date.

Latest

  • Wrapper-based Ensemble Feature Selection

    Find the most stable and predictive features using multiple learners and resampling techniques.

  • Time constraints in the mlr3 ecosystem

    Set time limits for learners, tuning and nested resampling.

  • Analyzing the Runtime Performance of tidymodels and mlr3

    Compare the runtime performance of tidymodels and mlr3.

  • Survival modeling in mlr3 using Bayesian Additive Regression Trees (BART)

    Demonstrate use of survival BART on the lung dataset via mlr3proba and distr6.

  • Spatial Data in the mlr3 Ecosystem

    Run a land cover classification of the city of Leipzig.

No matching items
    See all posts

    Train and Evaluate Models

    • Imbalanced Data Handling

      Handle imbalanced data with oversampling, undersampling, and SMOTE imbalance correction.

    • Resampling - Stratified, Blocked and Predefined

      Apply stratified, block and custom resampling.

    • Factor Encoding

      Encode factor variables in a task.

    • German Credit Series

      Train, tune and pipeline different machine learning algorithms.

    No matching items
    See all posts

    Optimize Models

    • Recursive Feature Elimination on the Sonar Data Set

      Utilize the built-in feature importance of models.

    • Hyperband Series

      Use the Hyperband optimizer with different budget parameters.

    • Practical Tuning Series

      Start with a tuned SVM and finish with a AutoML model.

    • Default Hyperparameter Configuration

      Run the default hyperparameter configuration of learners as a baseline.

    No matching items
    See all posts

    Build Pipelines

    • A Pipeline for the Titanic Data Set

      Create new features and impute missing values with a pipeline.

    • Pipelines, Selectors, Branches

      Build a preprocessing pipeline with branching.

    • Target Transformations via Pipelines

      Transform the target variable.

    • Tuning a Complex Graph

      Tune a preprocessing pipeline and multiple tuners at once.

    No matching items
    See all posts

    Apply Technical Tools and Run Special Tasks

    • Time constraints in the mlr3 ecosystem

      Set time limits for learners, tuning and nested resampling.

    • Production Example Using Plumber and Docker

      Write a REST API using plumber and deploy it using Docker.

    • Visualization in mlr3

      Quickly plot objects of the mlr3 ecosystem.

    • Spatial Data in the mlr3 Ecosystem

      Run a land cover classification of the city of Leipzig.

    No matching items
    See all posts

    Exercise Collection for Practice and Learning

    This curated exercise collection offers hands-on, solvable exercises focused on core concepts, from basic modeling and resampling to tuning and advanced techniques. Each exercise follows a consistent structure: clear learning goal, prerequisites, structured task list with hints how to solve them, collapsible password-protected solutions, and key take-aways designed to support systematic practice and self-paced learning.

    See all exercises

    Created with for #rstats.

    CC BY SA 4.0

    Website powered by quarto and Netlify.