Figure 13.

Structures of four types of ANNS: (a) multilayer perceptron, where x_i represents inputs, O_i represents output neurons, h_i represents hidden layer neurons, and w_i represents the weights between neurons; (b) wavelet, where Ψ represents the wavelet function, t represents the translation coefficient, λ represents the dilation coefficient; (c) radial basis function, where R_i represents the radial basis function; and (d) Elman, where u_i represents components in the hidden and undertake layers. Reprinted with permission from [133]. ©2019 Elsevier.