Title :
A Method to Determine the Parameters of Phase Space Reconstruction Based on the Neural Network
Author :
Liu, Runjie ; Hou, Zhan ; Shen, Jinyuan
Author_Institution :
Sch. of Inf. Eng., Zhengzhou Univ., Zhengzhou, China
Abstract :
In classic phase space reconstruction, the time lag is identical. In our research, the different time lags are found more effectively for teletraffic forecasting. In this paper, a method to determine the different time lags in phase space reconstruction is proposed. Simulation results show that the prediction is more accurate by using the different time lags in reconstruction phase space.
Keywords :
delays; neural nets; phase space methods; time series; neural network; phase space reconstruction parameters determination; teletraffic forecasting; time lag determination; Approximation algorithms; Chaos; Delay effects; Feedforward neural networks; Neural networks; Neurons; Predictive models; Resonance light scattering; Time series analysis; Vectors; Phase Space Reconstruction; TDNN; embedding dimensions; pruning algorithm; time lags;
Conference_Titel :
Chaos-Fractals Theories and Applications, 2009. IWCFTA '09. International Workshop on
Conference_Location :
Shenyang
Print_ISBN :
978-0-7695-3853-2
DOI :
10.1109/IWCFTA.2009.65
