Time series prediction model plays an important role in stock price prediction, such as ARIMA, LSTM neural network. However, due to the need for stationary assumption of time series itself and the problems of high dimension and high noise, the common time series prediction methods have limitations. Based on this, this paper propose a framework for the optimization of the stock price time series prediction model. The proposed method uses the intra-day price as the auxiliary variable and obtains the function feature information based on Karhunen-Loève expansion. Considering that the feature variables after dimension reduction still have problems such as information loss and irrelevant noise. This paper use data enhancement method to improve the effective information of feature variables and reduce the influence of irrelevant noise. Then, since the potential model structure between the characteristic variable and the residual sequence is unknown, this paper develop a weighted ensemble regression method based on information gain to balance the variance and deviation of the prediction model, thereby improving the prediction accuracy. The actual data analysis results show that the proposed method can greatly improve the fitting accuracy of ARIMA and LSTM neural networks for stock prices. Finally, the optimization framework can also be used for the prediction of average temperature, air quality and port cargo flow.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.