A combined approach for long-term series prediction: Renyi permutation entropy with BEA predictor filter

H. DANIEL PATIÑO; VICTOR H. SAUCHELLI; JULIÁN A. PUCHETA; GUSTAVO JUAREZ; CRISTIAN M. RODRÍGUEZ RIVERO; SERGIO O. LABORET

Buenos Aires

Congreso; IEEE ARGENCON 2016; 2016

Universidad Tecnológica Nacional - FR Buenos Aires

In order to predict long-term series, a Bayesianenhanced approach (BEA) combining permutation entropy(BEMA) is presented. The motivation of the proposed filter isto predict long-term time series by changing the structure ofthe predictor filter according to data model selected, thencomputational results are evaluated on high roughness timeseries selected from benchmark, in which they are comparedwith recent artificial neural networks (ANN) nonlinear filterssuch as Bayesian Enhanced approach (BEA) and BayesianApproach (BA). These results support the applicability ofpermutation entropy in analyzing the dynamic behavior ofchaotic time series for long-term series predictions.