Examinando por Autor "Cogollo, M.R."

Mostrando 1 - 2 de 2
  • No hay miniatura disponible
    Ítem
    Are neural networks able to forecast nonlinear time series with moving average components?
    (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2015-07-01) Cogollo, M.R.; Velásquez, J.D.; Universidad EAFIT. Escuela de Ciencias; Modelado Matemático
    In nonlinear time series forecasting, neural networks are interpreted as a nonlinear autoregressive models because they take as inputs the previous values of the time series. However, the use of neural networks to forecast nonlinear time series with moving components is an issue usually omitted in the literature. In this article, we investigate the use of traditional neural networks for forecasting nonlinear time series with moving average components and we demonstrate the necessity of formulating new neural networks to adequately forecast this class of time series. Experimentally we show that traditional neural networks are not able to capture all the behavior of nonlinear time series with moving average components, which leads them to have a low capacity of forecast. © 2015 IEEE.
  • No hay miniatura disponible
    Ítem
    Numerical-analytical solutions of predator-prey models
    (World Scientific and Engineering Academy and Society, 2013-01-01) González-Parra Gilberto, G.; Arenas, A.J.; Cogollo, M.R.; Universidad EAFIT. Escuela de Ciencias; Modelado Matemático
    This paper deals with the construction of piecewise analytic approximate solutions for nonlinear initial value problems modeled by a system of nonlinear ordinary differential equations. In real world several biological and environmental parameters in the predator-prey model vary in time. Thus, non-autonomous systems are important to be studied. We show the effectiveness of the method for autonomous and non-autonomous predator-prey systems. The method we have used is called the differential transformation method which has some suitable properties such as accuracy, low computational cost, easiness of implementation and simulation as well as preserving properties of the exact theoretical solution of the problem. The accuracy of the method is checked by numerical comparison with fourth-order Runge-Kutta results applied to several predator-prey examples.