Evaluation.log_loss: Evaluation.log_loss

Description Usage Arguments Value

View source: R/Evaluation.R

Description

Function to calculate the log loss evaluation metric.

Usage

1
Evaluation.log_loss(data.observed, data.predicted, eps = 1e-15)

Arguments

data.observed

the true data (Y).

data.predicted

the Y outcome from the estimator.

eps

is a small offset to let the log not go to Inf.

Value

double the calculated log loss.


frbl/OnlineSuperLearner documentation built on Feb. 9, 2020, 9:28 p.m.