Model
You are now setting up your model. Since you chose a penalized logistic regression, know by its friends as Lasso, you need to find out what is the best penalty value, and do it through a search algorithm.
The recipe you built to engineer your features prior to modeling is already loaded.
Deze oefening maakt deel uit van de cursus
Feature Engineering in R
Oefeninstructies
- Set up the penalty for tuning.
- Bundle your model and recipe in a workflow.
Praktische interactieve oefening
Probeer deze oefening eens door deze voorbeeldcode in te vullen.
# Set up the penalty for tuning
lr_model <- logistic_reg() %>% set_engine("glmnet") %>%
set_args(mixture = 1, penalty = ___)
lr_penalty_grid <- grid_regular(penalty(range = c(-3, 1)),levels = 30)
# Bundle your model and recipe in a workflow
lr_workflow <- workflow() %>%
___(lr_model) %>%
___(___)
lr_workflow