evaluation: Evaluation of prediction performance

In AlineTalhouk/splendid: SuPervised Learning ENsemble for Diagnostic IDentification

Evaluation of prediction performance

Description

Evaluation of prediction performance on the OOB set is done using various measure for classification problems.

Usage

evaluation(x, y, plot = FALSE)

Arguments

x	true class labels
y	predicted class labels
plot	logical; if TRUE a discrimination plot and reliability plot are shown for each class

Details

The currently supported evaluation measures include discriminatory measures like log loss, AUC, and PDI, macro-averaged PPV (Precision)/Sensitivity (Recall)/F1-score, accuracy (same as micro-averaged PPV Sensitivity/F1-score), Matthew's Correlation Coefficient (and its micro-averaged analog), Kappa, G-mean, and class-specific accuracy/PPV/NPV/Sensitivity/Specificity/F1-score/MCC/Kappa/G-mean.

Value

A list with one element per evaluation measure except for the cs element, which returns a list of class-specific evaluation measures.

Author(s)

Derek Chiu

Examples

data(hgsc)
class <- factor(attr(hgsc, "class.true"))
set.seed(1)
training.id <- sample(seq_along(class), replace = TRUE)
test.id <- which(!seq_along(class) %in% training.id)
mod <- classification(hgsc[training.id, ], class[training.id], "xgboost")
pred <- prediction(mod, hgsc, class, test.id)
evaluation(class[test.id], pred)

AlineTalhouk/splendid documentation built on Feb. 23, 2024, 9:37 p.m.