Neural networks with node gates
H.M. Myint, T. Murata, A. Nakazono, Kotaro Hirasawa · 2002
Function approximation problems for ordinary neural networks may be rather difficult, if the function becomes complicated, due to the necessity of big network size and the possibilities of many local minima. A promising way to solve these difficulties is the localization of the problem. According to this concept, a new architecture of a neural network is proposed namely neural network with node gates. In the paper, a function approximation example is provided to demonstrate the better performance of the proposed network than the ordinary neural network.