Output details
13 - Electrical and Electronic Engineering, Metallurgy and Materials
University of Manchester : B - Electrical and Electronic Engineering
Nonlinear system identification based on internal recurrent neural networks
This paper proposes a novel approach to nonlinear complex system identification based on internal recurrent neural networks (IRNN). The resulting modular neural model has been used by researchers in several fields seeking design control strategies for complex non-linear systems. In particular, the medical community has exploited the innovative principle of modularity and IRNN with a modified backpropagation algorithm; for example, Leon et al. (doi 10.1016/j.jfranklin.2012.02.011) use it for diabetes and Sankari and Adeli (doi 10.1016/j.jneumeth.2011.01.027) for the diagnosis of Alzheimer's disease and other neurological disorders (doi 10.1007/978-3-642-17569-5_1).