Simplification of a specific two-hidden-layer feedforward networks

A specific two-hidden-layer feedforward networks (TLFNs) proposed by G.B. Huang (2003) is presented in this paper. A method is introduced to simplify the structure of the TLFNs by introducing a new type of quantizers that unite two previous neurons A^(p) and B^(p) into a single neuron. Those new quantizers choose a special type of function as the neural network´s activation function, which leads to the new TLFNs with 2√((m+1)N) hidden neurons can learn N distinct samples (x_i, t_i) with negligibly small error, where m is the number of output neurons, and unlike Huang´s TLFNs require 2√((m+2)N) hidden neurons. Moreover, it is not necessary to estimate the quantizer value U defined in Huang´s TLFNs, which is fixed in our new model of TLFNs. It can reduce significantly and markedly the complexity and computation of neural networks.