Skip to main content
. 2020 Feb 5;15(2):e0228025. doi: 10.1371/journal.pone.0228025

Fig 5. Evaluation of prediction performance.

Fig 5

Results on the EPILEPSIAE dataset are shown in the upper row(A-C). Results on the public Epilepsyecosystem are shown in the middle row (D-F) and the results on the private Epilepsyecosystem dataset (benchmark) are shown in the lower row (G-I). Performance of each patient is represented by a circle, for accuracy (A, D, G), specificity-sensitivity plot (B, E, H) and negative and positive predictive value (C, F, I). Identical colors are used to represent each patient across all nine subplots. The hatched area represents results attainable by a random classifier.