pred_LB: Predict logistic biplot and thresholds by variable

In jgbabativam/BiplotML: Biplots Estimation with Algorithms ML

Predict logistic biplot and thresholds by variable

Description

Predicts the binary matrix and calculates the optimal thresholds per variable that minimize the Balanced Accuracy (BACC)

Usage

pred_LB(object, x, ncuts = 100)

Arguments

object	BiplotML object
x	Binary matrix.
ncuts	Number of equidistant cuts between 0 and 1 that will be evaluated. By default ncuts = 100

Details

The threshold for each variable is lowered to minimize the Balanced Error Rate (BER).

BER = 1 - \frac{1}{2} (\frac{TP}{TP+FN} + \frac{TN}{TN+FP}),

where TP is the number of true positives, TN is the number of true negatives, FP is the number of false positives and FN is the number of false negatives

Value

This function returns the thresholds per variable, the predicted matrix, the confusion matrix and the BER.

Examples

data("Methylation")
LB <- LogBip(Methylation, plot = FALSE)
out <- pred_LB(LB, Methylation)

jgbabativam/BiplotML documentation built on July 31, 2022, 11:10 a.m.