A Novel Approach for Parameter Estimation Of Discrete Dynamical Systems Using Self-Organizing Maps

In this paper, a novel approach towards parameter estimation of discrete dynamical systems with strange attractors is proposed which relies on parameter estimator self-organizing maps (PESOM). Theoretical justifications and a computational complexity analysis are presented regarding the efficiency of PESOM based cost function. Furthermore, experimental results demonstrating PESOM's superior performance compared to previously proposed cost functions such as Gaussian Mixture Model (GMM), Mean-Squared Error (MSE) and Return Map Fingerprint (RMF) are presented.