MeasureClassifCosts: Cost-sensitive Classification Measure

In mllg/mlr3: Machine Learning in R - Next Generation

Description

Uses a cost matrix to create a classification measure. True labels must be arranged in columns, predicted labels must be arranged in rows. The cost matrix is stored as slot $costs. Costs are aggregated with the mean.

Format

R6::R6Class() inheriting from MeasureClassif.

Construction

MeasureClassifCosts$new(costs = NULL, normalize = TRUE)
mlr_measures$get("classif.costs")
msr("classif.costs")

• costs :: matrix()
  Numeric matrix of costs (truth in columns, predicted response in rows).

• normalize :: logical(1)
  If TRUE, calculate the mean costs instead of the total costs.

See Also

Dictionary of Measures: mlr_measures

Examples

# get a cost sensitive task
task = tsk("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)

# mlr3 needs truth in columns, predictions in rows
costs = t(costs)

# create measure which calculates the absolute costs
m = msr("classif.costs", id = "german_credit_costs", costs = costs, normalize = FALSE)

# fit models and calculate costs
learner = lrn("classif.rpart")
rr = resample(task, learner, rsmp("cv", folds = 3))
rr$aggregate(m)

mllg/mlr3 documentation built on Sept. 27, 2019, 9:38 a.m.