Phase space reconstruction in sliver evenness series

According to the nonlinear characteristics of sliver evenness,a local linear forecasting method and a global neural networks forecasting method were given to predict a segment of sliver evenness. The results show that the two methods are good at forecasting the series in short-term,and the local linear forecasting is better than the other in this condition. At the time,it also approves the validity of the application of chaos nonlinear in the analysis to sliver evenness.