logloss: Negative log loss, also known as logarithmic loss or...
In chuvanan/metrics: Metrics For Evaluating Machine Learning Models

Description Usage Arguments Value Note Author(s) See Also Examples

Log loss is an effective metric for measuring the performance of a classification model where the prediction output is a probability value between 0 and 1.

Log loss quantifies the accuracy of a classifier by penalizing false classifications. A perfect model would have a log loss of 0. Log loss increases as the predicted probability diverges from the actual label. This is the cost function used in logistic regression and neural networks.

mtr_log_loss computes the elementwise log loss between two numeric vectors. While mtr_mean_log_loss computes the average log loss between two numeric vectors

mtr_log_loss(actual, predicted, eps = 1e-15)

mtr_cross_entropy(actual, predicted, eps = 1e-15)

mtr_mean_log_loss(actual, predicted, eps = 1e-15)

mtr_mean_cross_entropy(actual, predicted, eps = 1e-15)

`actual`	`[numeric]` Ground truth binary numeric vector containing 1 for the positive class and 0 for the negative class.
`predicted`	`[numeric]` A vector of estimated probabilities.
`eps`	`[numeric]` In case of predicted probability is equal zero or one, log loss is undefined, so probabilities are clipped to max(eps, min(1 - eps, p)). The default value of eps is 1e-15.

A numeric vector output

The logarithm used is the natural logarithm (base-e)

An Chu

mtr_brier_score

## log loss for scalar inputs, see how log loss is converging to zero
mtr_log_loss(1, 0.1)
mtr_log_loss(1, 0.5)
mtr_log_loss(1, 0.9)
mtr_log_loss(1, 1)

## sample data
act <- c(0, 1, 1, 0, 0)
pred <- c(0.12, 0.45, 0.9, 0.3, 0.4)

## log loss vector
mtr_log_loss(act, pred)
mtr_cross_entropy(act, pred)

## mean log loss
mtr_mean_log_loss(act, pred)
mtr_mean_cross_entropy(act, pred)