Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Cloud and Service-Oriented Computing
  4. Vol 2, Issue 1, 2022

An Optimization Framework for Stock Price Prediction Based on Statistical Information and Recursive Model Average -- Taking ARIMA Model as an Example

Download as PDF

DOI: 10.23977/csoc.2022.020103 | Downloads: 19 | Views: 1808

Author(s)

Jialan Xing 1, Yawen Li 2

Affiliation(s)

1 College of Arts and Sciences, Northeast Agricultural University, Harbin, Heilongjiang, 150006, China
2 School of Statistics, Capital University of Economics and Business, Beijing, 100070, China

Corresponding Author

Jialan Xing

ABSTRACT

Based on statistical information and recursive model average method, this paper proposes an optimization framework for stock price forecasting models. The proposed method uses intraday prices as auxiliary information and considers their functional and statistical characteristics. This framework continuously fits the residuals obtained from the original model prediction by a recursive model average method, weights the bias and variance of the prediction model, captures the functional statistical characteristics of intraday prices and the model structure among response variables, and finally optimizes the prediction accuracy of the original model. This framework is model free in theory. Taking the optimized ARIMA model as an example, the data analysis results show that the proposed method has better fitting performance and is robust compared to the ARIMA model. In addition, the proposed method can be extended in application scenarios such as average temperature prediction, traffic flow monitoring, and port cargo capacity prediction.

KEYWORDS

stock price forecasting, statistical information quantity, recursive model average method, functional characteristics, ARIMA

CITE THIS PAPER

Jialan Xing, Yawen Li, An Optimization Framework for Stock Price Prediction Based on Statistical Information and Recursive Model Average -- Taking ARIMA Model as an Example. Cloud and Service-Oriented Computing (2022) Vol. 2: 21-27. DOI: http://dx.doi.org/10.23977/csoc.2022.020103.

REFERENCES

[1] Yang Haimin, Pan Zhisong, Bai Wei. A review of time series forecasting methods. Computer Science, 2019, 46 (01): 21-28. 
[2] Wu Rongliang. An empirical analysis of the return rate of Shanghai A-share index based on AR model. Quotient, 2016 (03): 173-174. 
[3] Wu Yuxia, Wen Xin. Short-term stock price prediction based on ARIMA model. Statistics and Decision, 2016, 13 (23): 83-86. 
[4] Zhao Yi. An ARIMA and LSTM based combined forecasting model for civil aviation passenger reservation [J]. Computers and Modernization, 2020 (11): 65-69 + 76. 
[5] Meng Yi. Application of time series ARIMA and BP neural network combination model in CPI prediction. Journal of Shandong Agricultural University (Natural Science Edition), 2018, 49 (06): 1079-1083. 
[6] Tu Xiaomeng, Chen Qiangguo. Crime time series prediction based on ARIMA-LSSVM hybrid model [J]. Electronic Technology Application, 2015, 41 (02): 160-162 + 166.
[7] Wang Qingrong. Research and application of functional principal component analysis and functional linear regression model. Chongqing Technology and Business University, 2020. 
[8] Tang Yi, Feng Changhuan. Research on urban population in China based on functional principal component analysis. Journal of Yili Normal University (Natural Science Edition), 2019, 13 (03): 9-16. 
[9] Wu Fei and Chen Dirong. Functional data classification based on functional principal components [J]. Journal of Wuhan Textile University, 2019, 32 (02): 48-56.

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 219
Visits: 18558

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.