We investigate in detail the effects of noise on the performance of reservoir computing. We focus on an application in which reservoir computers are used to learn the relationship between different state variables of a chaotic system. We recognize that noise can affect differently the training and the testing phases. We find that the best performance of the reservoir is achieved when the strength of the noise that affects the input signal in the training phase equals the strength of the noise that affects the input signal in the testing phase. For all the cases we examined, we found that a good remedy to noise is to low-pass filter the input and the training/testing signals; this typically preserves the performance of the reservoir, while reducing the undesired effects of noise.

Biography

Francesco Sorrentino is full professor of Automatic Controls in the Department of Mechanical Engineering at the University of New Mexico. He received a master's degree in Industrial Engineering from the University of Naples Federico II (Italy) in 2003 and a Ph.D. in Control Engineering from the University of Naples Federico II (Italy) in 2007.

His expertise is in dynamical systems and controls, with particular emphasis on nonlinear dynamics and optimal control. His work includes studies on dynamics and control of complex dynamical networks, adaptation in complex systems, sensor adaptive networks, and the dynamics of reservoir computers in machine learning. He is interested in applying the theory of dynamical systems to model, analyze, and control the dynamics of complex distributed energy systems, such as power networks and smart grids. Subjects of current investigation are evolutionary game theory on networks (evolutionary graph theory), the dynamics of large networks of coupled neurons, and the use of optimal control to design drug dosage schedules for biomedical applications.

He has published more than 70 papers in international scientific peer reviewed journals and serves as an Associate Editor of the IEEE Control Systems Letters (L-CSS). His research is funded by the National Science Foundation, the Office of Naval Research, and the Defense Threat Reduction Agency.