The 3-input Euler polynomial neuronet (3IEPN) with weights-and-structure-determination (WASD) algorithm

Based on the function approximation theory and the Weierstrass approximation theorem, a novel 3-input Euler polynomial neuronet (3IEPN) is established in this paper to fit the 4-dimensional data (i.e., data in the form of 3 inputs and 1 output). In order to achieve satisfactory performance and efficacy of the 3IEPN, a weights-and-structure-determination (WASD) algorithm, which is based on the weights-direct-determination (WDD) method, is built up for the established 3IEPN. Numerical results further substantiate the superior performance of the proposed 3IEPN equipped with the WASD algorithm in terms of training, validation and prediction.