mlr3 tutorials

Over the past few months, I have written quite a few blog posts related to mlr3, so here is an overview of the different topics, simplest to most complex.

• Generalization to new subsets in R and Python explains how to code SOAK from first principles.
• Comparing ML Frameworks discusses advantages of mlr3 relative to a basic for loop, and the more recent tidymodels framework in R.
• New code for various kinds of cross-validation explains how to use mlr3resampling to implement SOAK, and variable sized train sets.
• Mammouth tutorial explains how to use a SLURM computer cluster for massive speedups of machine learning benchmark experiments.
• Cross-validation experiments with torch learners explains how to compare linear models in torch with other learners from outside torch like glmnet.
• Comparing neural network architectures using mlr3torch explains how to implement different neural network architectures, and make figures to compare their subtrain/validation/test error rates.
• Torch learning with binary classification explains how to implement a custom loss function for binary classification in mlr3torch.
• TODO lazy tensor blog, show memory advantages.
• Toulouse RUG slides explain advantages of torch and mlr3.