Skip to main content
Fig. 1 | Brain Informatics

Fig. 1

From: Statistical independence for the evaluation of classifier-based diagnosis

Fig. 1

Two simulated examples of confusion matrices, true class labels on the rows and predicted class labels on the columns. Both confusion matrices have the same prediction accuracy (PA), i.e. \(PA = 1 - \hat{\epsilon } = \frac{0+90}{100} = \frac{80+10}{100} = 90\,\%\). Nevertheless, in the first case, there is no evidence that the classifier is able to discriminate H from P, while in the second one there is.

Back to article page