Increasing prediction accuracy of pathogenic staging by sample augmentation with a GAN
Fig 4
Classification accuracy for randomly sampled data.
100G, 50G, 30G and 10G indicate classification accuracies using 5 times augmented data from 100%, 50%, 30%, and 10% randomly selected samples, respectively. 100O, 50O, 30O, and 10O indicate classification accuracies using 100%, 50%, 30%, and 10% randomly selected samples (same as those for 100G/50G/30G/10G), respectively. Classification algorithm used is 1DCNN, and each random sampling was performed 10 times.