loglossBinary {lares}

Logarithmic Loss Function for Binary Models

Description

This function calculates log loss/cross-entropy loss for binary models. NOTE: when result is 0.69315, the classification is neutral; it assigns equal probability to both classes.

Usage

loglossBinary(tag, score, eps = 0.001)

Arguments

tag

Vector. Real known label

score

Vector. Predicted value or model's result

eps

Numeric. Epsilon value

See Also

Other Model metrics: ROC(), conf_mat(), errors(), gain_lift(), model_metrics()