Proceedings of the 5th International Conference on E-Commerce and Internet Technology, ECIT 2024, March 15–17, 2024, Changsha, China

Research Article

Stock price prediction based on particle swarm algorithm optimised SVM univariate time series algorithm

Download34 downloads
  • @INPROCEEDINGS{10.4108/eai.15-3-2024.2346191,
        author={Yuhang  Li},
        title={Stock price prediction based on particle swarm algorithm optimised SVM univariate time series algorithm},
        proceedings={Proceedings of the 5th International Conference on E-Commerce and Internet Technology, ECIT 2024, March 15--17, 2024, Changsha, China},
        publisher={EAI},
        proceedings_a={ECIT},
        year={2024},
        month={5},
        keywords={svm stock price prediction time series},
        doi={10.4108/eai.15-3-2024.2346191}
    }
    
  • Yuhang Li
    Year: 2024
    Stock price prediction based on particle swarm algorithm optimised SVM univariate time series algorithm
    ECIT
    EAI
    DOI: 10.4108/eai.15-3-2024.2346191
Yuhang Li1,*
  • 1: Northeastern University
*Contact email: lyh15204130550@qq.com

Abstract

In this paper, using the Coca-Cola stock price data from the UCL public dataset, the support vector machine univariate time series algorithm optimised by particle swarm algorithm is used to conduct a prediction study on the stock opening price. By observing the X-Y scatter plot of the stock price and the actual stock price, it is found that the distribution of data points in the training set and the test set mainly focuses on the Y=X straight line, indicating that the predicted value is very close to the actual value and the model has good prediction effect. Further analysis of the line graphs of predicted and actual stock prices shows that the root mean square error (RMSE) of the training set is 0.78195 and the RMSE of the test set is 0.56134, which verifies the high accuracy of the model. Overall, the support vector machine univariate time series algorithm optimised based on particle swarm algorithm can not only accurately predict the stock price values, but also better capture the stock price trend