This paper proposes a traffic anomaly detection method based on mislearned autoencoder. By intentionally mislearning the patterns of abnormal samples, the method increases the reconstruction loss of ...