Classification models for the multivariate time series have gained
significant importance in the research community, but not much research has
been done on generating adversarial samples for these models. Such samples of
adversaries could become a security concern. In this paper, we pr