Published 2024-11-30
How to Cite
This work is licensed under a Creative Commons Attribution 4.0 International License.
Abstract
The stock market serves as a critical indicator of a nation's economic vitality, reflecting trends and potential risks. With increasing participation from investors seeking high returns, the demand for efficient and accurate stock price prediction methods has intensified. Traditional time-series statistical models such as ARMA, GARCH, and Markov models have been widely applied, yet limitations remain in their predictive performance. This study identifies the insufficient generalization capability of the Support Vector Regression (SVR) model in stock price forecasting and introduces an integrated learning algorithm to address this issue. By combining SVR with the Linear Regression (LR) and K-Nearest Neighbor (KNN) models, the proposed approach capitalizes on the strengths of each method. Experimental results across multiple datasets demonstrate that the integrated model outperforms the standalone SVR model in accuracy and robustness. Furthermore, this paper highlights future research opportunities, including incorporating additional influencing factors and optimizing SVR parameters, to further enhance predictive capabilities.