Uses a cost matrix to create a classification measure. The cost matrix is stored as slot "costs". Costs are aggregated with the mean.
MeasureClassifCosts$new(costs = NULL, normalize = TRUE)
Identifier for the measure.
Numeric matrix of costs (truth in columns, predicted response in rows).
TRUE, calculate the mean costs instead of the total costs.
# get a cost sensitive task task = mlr_tasks$get("german_credit") # cost matrix as given on the UCI page of the german credit data set # https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data) costs = matrix(c(0, 5, 1, 0), nrow = 2) dimnames(costs) = list(truth = task$class_names, predicted = task$class_names) print(costs)#> predicted #> truth good bad #> good 0 1 #> bad 5 0# mlr3 needs truth in columns, predictions in rows costs = t(costs) # create measure which calculates the absolute costs m = MeasureClassifCosts$new(id = "german_credit_costs", costs, normalize = FALSE) # fit models and calculate the costs resample(task, "classif.rpart", "cv3", measure = m)#> INFO [mlr3] Running learner 'classif.rpart' on task 'german_credit' (iteration 1/3)' #> INFO [mlr3] Running learner 'classif.rpart' on task 'german_credit' (iteration 2/3)' #> INFO [mlr3] Running learner 'classif.rpart' on task 'german_credit' (iteration 3/3)'#> <ResampleResult> of learner 'german_credit' on task 'classif.rpart' with 3 iterations #> Measure Min. 1st Qu. Median Mean 3rd Qu. Max. Sd #> german_credit_costs 270 288.5 307 320.7 346 385 58.71