When evaluating machine learning algorithms through resampling, it is preferable that each train/test partition will be a representative subset of the whole data set. This post covers three ways to achieve such reliable resampling procedures.
Post moved to mlr-org.com/gallery/2020-03-30-stratification-blocking/.
For attribution, please cite this work as
Dragicevic & Casalicchio (2020, March 30). mlr3gallery: Resampling - Stratified, Blocked and Predefined. Retrieved from https://mlr3gallery.mlr-org.com/posts/2020-03-30-stratification-blocking/
BibTeX citation
@misc{dragicevic2020resampling, author = {Dragicevic, Milan and Casalicchio, Giuseppe}, title = {mlr3gallery: Resampling - Stratified, Blocked and Predefined}, url = {https://mlr3gallery.mlr-org.com/posts/2020-03-30-stratification-blocking/}, year = {2020} }