A discussion on ‘Detection of intrusions in information systems by sequential change-point methods’ by Tartakovsky, Rozovskii, Blažek, and Kim

The focus of the paper under discussion is on the case when the attack appears only in a single channel. The authors recognize that this is a theoretical idealization of real life applications.I would like to expand on this using as illustration the Gaussian distributions with fixed, known covariance matrix. Depending on the specified out-of-control situation, one can distinguish between different changes in the mean vector and the control charts should be defined and used according to the various levels of the available a priori information about these changes.