TITLE:
Comparative Study on Forecasting Model for Stock Index Future Price
AUTHORS:
Wenbo Wang, He Nie
KEYWORDS:
Forecasting Model, Stock Index Future Price, Wavelet Analysis, PSO, Neural Networks
JOURNAL NAME:
Modern Economy,
Vol.9 No.4,
April
26,
2018
ABSTRACT: The wavelet neural network has been widely applied
in the forecasting field. However, due to its disadvantages of local minimal
and slow convergence speed, its forecasting precision is limited in some
extent. A forecasting model for stock index future price based on wavelet
analysis and the improved PSO-based neural network would be put forward in this
essay using the improved PSO-based algorithm to optimize the neural network. By
using the Shanghai and Shenzhen 300 stock index futures’ statistics as the
samples, the sym 8 wavelet
transformation techniques to denoise the statistics and then use the denoised
stats to train and test the improved PSO-based neural networks. The study results
show that the model
in this essay could efficiently improve the forecasting quality of the stock
index futures price.