Abstract: This paper presents an interesting methodology for a class of nonlinear system identification via the recurrent Extreme Learning Machine (ELM). The recurrent ELM has been used for training single hidden layer feed-forward neural networks. Vis-à-vis to various feed-forward NN the ELM is remarkably efficient and tends to reach global optimum with high convergence speed. For many real applications, in control system of view point can be represented by nonlinear autoregressive with external input (NARX) models. The recurrent ELM presents efficient tools for working with dynamical recurrent data. Recurrent Neural Network (RNN) is an adequate structure for NARX models and can be trained in online context.......
Key Word: adaptive identification, extreme learning machine, recurrent neural network, process modeling.
[1]. Horvath G. Neural networks in systems identification. In: Ablameyko S, Goras L, Gori M, Piuri (szerk.) V, editors.Neural Networks
in Measurement Systems. Amsterdam: IOS Press, 2003. pp. 43-78.
[2]. Narendra KS, Parthasaraty K.Neural networks and dynamical systems. J Approx Reason 1992;6:109-131.
[3]. O. Taouali, I. Elaissi, and H. Messaoud 2012 Design and comparative study of online method identification of nonlinear system in
RKHS space. Artificial Intelligence Review 2012; 4-37:289-300
[4]. Wang L, Zeng Y, Chen T. Back propagation neural network with adaptive differential evolution algorithm for time series
forecasting. J Expert syst Appl 2015; 42: 855-863.
[5]. G.-B. Huang, L. Chen, and C.-K. Siew, "Universal approximation using incremental constructive feedforward networks with
random hidden nodes," IEEE Transactions on Neural Networks, 2006; 7- 4 : 879–892.
