A-share market prediction and trading strategies / Lu Tianfeng
As a significant financial instrument, stocks have consistently attracted investors seeking profitable opportunities. Yet, forecasting stock prices remains challenging due to intricate market dynamics characterized by noise, nonlinearity, and temporal variability. Recent global crises ranging from p...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2024
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/15987/1/Lu_Tianfeng.pdf http://studentsrepo.um.edu.my/15987/2/Lu_Tianfeng.pdf http://studentsrepo.um.edu.my/15987/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | As a significant financial instrument, stocks have consistently attracted investors seeking profitable opportunities. Yet, forecasting stock prices remains challenging due to intricate market dynamics characterized by noise, nonlinearity, and temporal variability. Recent global crises ranging from pandemics to geopolitical tensions have heightened market volatility, underscoring the need for more robust predictive models. The rapid development of artificial intelligence and machine learning techniques, with their enhanced capacity to model complex nonlinear relationships, has rendered them increasingly essential in stock price prediction tasks. This study integrates Particle Swarm Optimization (PSO) with Long Short-Term Memory (LSTM) neural networks to improve predictive accuracy in the Chinese A-share market. Through a PSO-driven hyperparameter tuning process, we refine the LSTM architecture, enabling it to better capture intricate temporal dependencies and market patterns. Empirical results show that the PSO-LSTM model outperforms traditional LSTM, MLP neural networks, and conventional benchmark models in terms of key accuracy metrics (MSE, MAE, RMSE, MAPE, and |
|---|