i2mlr course Use Case: Bike Sharing Demand is ready for review now.
All .rds files are shrunken to data.tables if possible! (Note that tuned ranger is necessary for prediction.)
I'am aware that the handling of paramset and trafo does not seem very straight foreward. Especially the hyperparameter names have long unintuitive names. This inconvenience is made up by the neccesity to wrap the pipe in a graph before branching...
Is there any handy way to solve this problem?
i2mlr course Use Case: Bike Sharing Demand is ready for review now. All
.rds
files are shrunken to data.tables if possible! (Note that tuned ranger is necessary for prediction.)I'am aware that the handling of paramset and trafo does not seem very straight foreward. Especially the hyperparameter names have long unintuitive names. This inconvenience is made up by the neccesity to wrap the pipe in a graph before branching... Is there any handy way to solve this problem?