Analysis of Multi-Channel Subdural EEG by Recurrent Neural Networks

State Estimation of Discrete-Time Switched Neural Networks With Multiple Communication Channels

“…A universal way for handling the random measurement losses over a single channel is to utilize the Bernoulli distributed white sequence specified by a conditional probability distribution in the measurement description (see [28]- [31]). In practical neural networks, however, more than one channel is highly likely simultaneously available, which will reduce the influence of measurement losses and even the disconnection of the single channel in the transmission network, see the applications in neural networks [32], [33]. Therefore, it will be of great necessity and significance to consider multiple communication channels in the studies of diverse issues for improving the system performance.…”

“…Hence, the estimator gains with optimal generalized H 2 performance can be determined by solving the following convex optimization problem: min γ s.t. (30)-(32).…”