Reconstruction-based anomaly detection for multivariate time series using contrastive generative adversarial networks

Jiawei Miao, Haicheng Tao, Haoran Xie, Jianshan Sun, Jie Cao

Abstract: Highlights•Data augmentation with geometric mask is effective in mitigating overfitting.•Reconstructed data can be treated as positive samples for the contrastive constraint.•Contrastive learning is seamless integrated into the GAN framework.