Stochastic Composite Neural Network For Financial Time Series Prediction And Statistical Analysis

Due to the frequent and violent fluctuation of energy futures prices,the investment risk of energy investors is increased.Forecasting energy futures prices and fluctuations has progressively become the focus of research.However,traditional prediction model only conducts forecasting based on historical data without considering the behavior of the market,resulting in poor accuracy.In this paper,the random time effective function that considers the timeliness of historical data and the random change of market environment is applied to different neural network prediction models to build hybrid models with higher prediction accuracy.Because of the non-linearity of financial time series,the ensemble empirical mode decomposition can suppress pattern confusion and restore signal essence in effect.This paper combines EEMD into wavelet neural network with random time effective(WNNRT)to establish a hybrid neural network prediction model,which is called EEMD-WNNRT,and multiscale complexity invariant distance(MCID)is utilized to evaluate the predicting performance of EEMD-WNNRT model.Furthermore,the proposed model is compared with the traditional models in predicting the impact of global energy prices using other error evaluations,and its corresponding superiority is proved.Furthermore,long short term memory model(LSTM)has the characteristics of selective memory of time series,which is very suitable for the prediction of price time series.In this paper,the random time effective function,which can give different weights to historical data,is applied to the long short term memory model to establish a novel prediction model that has higher prediction accuracy.This model has the effect of random movement and keeps the trend fluctuation of the original nonlinear data,which makes the prediction more accurate and more credible.And using multiscale crosssample entropy(MCSE)as an innovative method to reveal the performance of prediction.Moreover,comparing with other models selected in this paper to demonstrate the advantages and superiority of the proposed model by other error evaluations.