Related Experiment Video
Updated: Apr 15, 2026
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
16.5K
A stock market forecasting model combining two-directional two-dimensional principal component analysis and radial
Zhiqiang Guo1, Huaiqing Wang2, Jie Yang1
1Key Laboratory of Fiber Optic Sensing Technology and Information Processing, School of Information Engineering, Wuhan University of Technology, Wuhan, China.
Plos One
|April 8, 2015
Summary
This study introduces a hybrid model using two-directional two-dimensional principal component analysis ((2D)2PCA) and a Radial Basis Function Neural Network (RBFNN) for accurate stock market forecasting. The novel approach enhances prediction accuracy compared to traditional methods.
Area of Science:
- Computational Finance
- Machine Learning
- Time Series Analysis
Background:
- Accurate stock market forecasting is crucial for financial decision-making.
- Traditional forecasting models often struggle with high-dimensional financial data.
- Dimensionality reduction techniques are essential for improving model efficiency and performance.
Purpose of the Study:
- To propose and implement a hybrid model for stock market behavior forecasting.
- To leverage the strengths of (2D)2PCA for feature extraction and RBFNN for prediction.
- To evaluate the model's effectiveness on the Shanghai stock market index.
Main Methods:
- A hybrid model combining two-directional two-dimensional principal component analysis ((2D)2PCA) and a Radial Basis Function Neural Network (RBFNN).
- Selection of 36 technical stock market variables as input features.
- Application of a sliding window technique for data preparation.
- (2D)2PCA for data dimensionality reduction and intrinsic feature extraction.
- RBFNN for forecasting next-day stock prices or movements.
Main Results:
- The proposed hybrid model demonstrated a good level of fitness in forecasting stock market behavior.
- Empirical results showed the model's superiority over traditional Principal Component Analysis (PCA) and Independent Component Analysis (ICA) based models.
- The model also outperformed alternative models utilizing Multilayer Perceptron (MLP).
Conclusions:
- The hybrid (2D)2PCA-RBFNN model offers a robust and effective approach to stock market forecasting.
- The integration of advanced dimensionality reduction with neural networks enhances predictive accuracy.
- This method provides a valuable tool for financial market analysis and prediction.
