Imbalanced Fault Classification of Bearing via Wasserstein Generative Adversarial Networks with Gradient Penalty

<div>The extracted features of the generator for OF3 samples: (a) the first hidden layer, (b) the second hidden layer, (c) the output layer, and (d) the raw OF3 samples.</div>

Shock and Vibration

fig4

Figure 4

Figure 4: Imbalanced Fault Classification of Bearing via Wasserstein Generative Adversarial Networks with Gradient Penalty 