Short-term stock market price trend prediction using a comprehensive deep learning system

Area of Science:

• * Computational Finance
• * Data Science
• * Machine Learning

Background:

• * The increasing volume of big data has amplified the need for sophisticated stock market prediction models.
• * Deep learning techniques are gaining prominence for analyzing complex financial market trends.
• * Existing models often lack comprehensive feature engineering tailored to specific market dynamics.

Purpose of the Study:

• * To develop and evaluate a customized deep learning model for predicting Chinese stock market price trends.
• * To investigate the impact of comprehensive feature engineering and data pre-processing on prediction accuracy.
• * To contribute novel insights into stock analysis for both financial and technical research communities.

Main Methods:

• * Collection and pre-processing of two years of Chinese stock market data.
• * Application of multiple advanced feature engineering techniques.
• * Development of a customized deep learning system for trend prediction.
• * Comparative evaluation against standard machine learning models.

Main Results:

• * The proposed deep learning model, enhanced by comprehensive feature engineering, demonstrated superior performance.
• * The system achieved high overall accuracy in predicting stock market price trends.
• * Detailed analysis confirmed the effectiveness of specific prediction term lengths and pre-processing methods.

Conclusions:

• * Comprehensive feature engineering is critical for enhancing the accuracy of deep learning-based stock market prediction.
• * The customized deep learning system offers a robust solution for stock trend forecasting.
• * This research provides valuable contributions to the fields of financial big data analysis and machine learning applications.