mlr: Machine Learning in R

By Bernd Bischl, Michel Lang, Lars Kotthoff, Julia Schiffner, Jakob Richter, Erich Studerus, Giuseppe Casalicchio, Zachary M. Jones. 2016

676 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

To tune or not to tune the number of trees in random forest? (2017, methodology): “om forest constructed with the R package randomForest (Liaw and Wiener, 2002) or ranger (Wright, 2016) as input and can calculate the OOB curve for any measure that is available from the mlr package (Bischl et al., 2016). See our R-code for application examples. The OOBCurve package will soon be ava”
Automated Algorithm Selection: Survey and Perspectives (2018, methodology): “…state-of-the-art methods, which are of special signiﬁcance in this context, are facilitated by benchmarking platforms and libraries, such as ASL IB (Bischl et al., 2016a), ACL IB (Hutter et al., 2014a) and HPO LIB (Eggensperger et al., 2013) for algorithm selection, conﬁguration and…”
DALEX: explainers for complex predictive models (2018, methodology): “…DALEX supports models generated with the most popular frameworks for classification or regression, such as caret (from Jed Wing et al., 2016), mlr (Bischl et al., 2016), Random Forest (Liaw and Wiener, 2002), Gradient Boosting Machine (Ridgeway, 2017) and Generalized Linear Models (Dobson, 1990).”
OpenML: An R package to connect to the machine learning platform OpenML (2017, methodology): “Second, we brieﬂy introduce the mlr package (Bischl et al, 2016; Schiﬀner et al, 2016), which is a machine learning toolbox for R (R Core Team, 2016) and oﬀers a uniﬁed interface to many machine learning algorithms.”
Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges (2021, methodology): “In the related problem of algorithm selection (Rice, 1976), we usually train a meta-learning model offline to select an optimal element from a finite set of algorithms or HPCs (Bischl et al., 2016).”
Multi-Objective Counterfactual Explanations (2020, methodology): “For all other hyperparameters of the models, the default values of the mlr (version 2.16.0) and keras (version 2.2.5.0) R packages were chosen [3,1].”
Random forest versus logistic regression: a large-scale benchmark experiment (2018, methodology): “10) for higher abstraction and a simpler way to conduct benchmark studies [26], OpenML (version 1.”
mlr3: A modern object-oriented machine learning framework in R (2019, background): “The package is a complete reimplementation of the mlr (Bischl et al., 2016) package that leverages many years of experience and learned best practices to provide a state-of-the-art system that is powerful, flexible, extensible, and maintainable.”
OpenML Benchmarking Suites (2017, methodology): “…environments such as the OpenAI Gym [Brockman et al., 2016] run and evaluate reinforcement learning experiments, the COCO suite standardizes benchmarking for black-box optimization [Hansen et al., 2020] and ASLib provides a benchmarking protocol for algorithm selection [Bischl et al., 2016a].”
Tunability: Importance of Hyperparameters of Machine Learning Algorithms (2018, methodology): “All our experiments are executed in R and are run through a combination of custom code from our random bot Kühn et al. (2018), the OpenML R package (Casalicchio et al., 2017), mlr (Bischl et al., 2016) and batchtools (Lang et al., 2017) for parallelization.”
Hyperparameters and tuning strategies for random forest (2018, methodology): “It can be changed to any of the 50 measures currently implemented in the R package mlr and documented in the online tutorial (Schiﬀner et al., 2016): http://mlr-org.github.io/mlr-tutorial/devel/html/measures/index.html • The ﬁnal recommended hyperparameter setting is calculated by taking the best 5…”