Re-calibrate classification probability predictions — adjust_probability

Calibration is the process of adjusting a model's outputted probabilities to match the observed frequencies of events. This technique aims to ensure that when a model predicts a certain probability for an outcome, that probability accurately reflects the true likelihood of that outcome occurring.

Usage

adjust_probability_calibration(x, method = NULL)

Arguments

x: A tailor().
method: Character. One of "logistic", "multinomial", "beta", "isotonic", or "isotonic_boot", corresponding to the function from the probably package probably::cal_estimate_logistic(), probably::cal_estimate_multinomial(), etc., respectively.

Data Usage

This adjustment requires estimation and, as such, different subsets of data should be used to train it and evaluate its predictions. See the section by the same name in ?workflows::add_tailor() for more information on preventing data leakage with postprocessors that require estimation. When situated in a workflow, tailors will automatically be estimated with appropriate subsets of data.

Examples

if (FALSE) {
library(modeldata)

# split example data
set.seed(1)
in_rows <- sample(c(TRUE, FALSE), nrow(two_class_example), replace = TRUE)
d_potato <- two_class_example[in_rows, ]
d_test <- two_class_example[!in_rows, ]

head(d_potato)

# specify calibration
tlr <-
  tailor() %>%
  adjust_probability_calibration(method = "logistic")

# train tailor on a subset of data. situate in a modeling workflow with
# `workflows::add_tailor()` to avoid having to specify column names manually
tlr_fit <- fit(
  tlr,
  d_potato,
  outcome = c(truth),
  estimate = c(predicted),
  probabilities = c(Class1, Class2)
)

# apply to predictions on another subset of data
head(d_test)

predict(tlr_fit, d_test)
}