Supervised identification of non linear models using moments and non-Gaussian signals

This work concerns the problem of the supervised identification of the parameters of non linear models using 3rd order moments. The input sequence is assumed to be independent and identically distributed (i.i.d), zero mean and must be non-Gaussian. The developed algorithm is tested and compared with other method developed in literature. Simulation examples are provided to verify the performance of the developed algorithm. The obtained results demonstrate the efficiency and the accuracy of the developed algorithm for non linear model identification under various sample sizes.