MulticlassClassificationMetrics.LogLoss Property
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Gets the average log-loss of the classifier. Log-loss measures the performance of a classifier with respect to how much the predicted probabilities diverge from the true class label. Lower log-loss indicates a better model. A perfect model, which predicts a probability of 1 for the true class, will have a log-loss of 0.
public double LogLoss { get; }member this.LogLoss : doublePublic ReadOnly Property LogLoss As DoubleProperty Value
Remarks
The log-loss metric is computed as follows: $LogLoss = - \frac{1}{m} \sum_{i = 1}^m log(p_i)$, where $m$ is the number of instances in the test set and $p_i$ is the probability returned by the classifier of the instance belonging to the true class.