Multi-Step-Ahead Stock Index Prediction Based on a Novel Hybrid Model

Accurate prediction of stock prices is crucial both for government policy-making and for market participants to make investment decisions. This paper proposes an intelligent financial time series forecasting model. The model integrates Ensemble Empirical Mode Decomposition (EEMD), Least Squares Supp...

Full description

Saved in:
Bibliographic Details
Published in2023 8th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA) pp. 185 - 189
Main Authors Zhang, Chengzhao, Shi, Weimei, Tang, Huiyue, Guo, Fanyong
Format Conference Proceeding
LanguageEnglish
Published IEEE 26.04.2023
Subjects
Adaptation models
EEMD
Feature extraction
Indexes
Kernel
KNN
LSSVR
prediction
Predictive models
Solid modeling
Time series analysis
Online AccessGet full text

Cover

More Information
Summary:Accurate prediction of stock prices is crucial both for government policy-making and for market participants to make investment decisions. This paper proposes an intelligent financial time series forecasting model. The model integrates Ensemble Empirical Mode Decomposition (EEMD), Least Squares Support Vector Regression (LSSVR) and weighted k-nearest neighbor (KNN), and proposes a new decomposition-cluster-ensemble stock price prediction method. This model is innovative in the construction of the feature extraction process, and proposes an improved algorithm that is more effective than the simple KNN prediction. The empirical results confirm the prediction effect of the model on Chinese stock index.
ISSN:2832-3734
DOI:10.1109/ICCCBDA56900.2023.10154678