In this study, we propose a DB-CGAN model to deal with distribution bias issues for imbalanced deep learning in industrial IoT. An integrated data augmentation framework is constructed based on introduction of a complementary classifier into the basic GAN model. Specifically, a conditional generator with random labels is designed and trained adversarially with the classifier to effectively enhance the augmentation of number of data samples in minority classes, while a weight sharing scheme is newly devised between two separated feature extractors, enabling the collaborative adversarial training among generator, discriminator, and classifier. An augmentation algorithm is then developed to facilitate intelligent anomaly detection, which can efficiently improve the classification accuracy based on the correction of distribution bias. Experiment evaluations based on two real-world imbalanced datasets demonstrate the outstanding performance of our proposed model in tackling distribution bias issues for imbalanced learning in industrial IoT applications, compared with five baseline methods.
ASJC Scopus subject areas
- コンピュータ サイエンスの応用