基于深度学习技术的股价预测及量化交易策略探讨
Stock Price Prediction and Quantitative Trading Strategy Based on Deep Learning

近几年来，卷积神经网络(CNN)、长短期记忆网络(LSTM)、门控循环单元(GRU)和注意力机制模型等深度学习技术，在金融科技领域尤其是股价预测和量化交易策略的制定上，已经成为一个非常活跃的研究领域。论文通过分析上证50指数及其成分股的数据，验证了BiGRU-CNN-Attention模型在预测准确性上的优势；考虑到不同投资者的风险承受能力和收益预期，论文设计了保守型、稳健型和极端激进型三种不同风险偏好的投资策略，揭示了在风险和回报之间平衡的效果。结果表明，结合深度学习模型预测和适当的投资策略，不仅可以有效提升投资组合的性能，还可以为投资者提供了定制化的投资方案，进一步凸显深度学习技术在金融市场决策中的应用潜力。
In recent years, deep learning technologies such as Convolutional Neural Networks (CNN), Long Short Term Memory Networks (LSTM), Gated Recurrent Units (GRU), and Attention Mechanism Models have become an active research field in the field of financial technology, especially in stock price prediction and quantitative trading strategy formulation. The paper verifies the advantage of the BiGRU-CNN Attention model in prediction accuracy by analyzing the data of the Shanghai Stock Exchange 50 Index and its constituent stocks; considering the risk tolerance and return expectations of different investors, the paper designs three investment strategies with different risk preferences: conservative, robust, and extremely aggressive, revealing the effect of balancing risk and return. The results indicate that combining deep learning models with appropriate investment strategies can not only effectively improve the performance of investment portfolios, but also provide customized investment plans for investors, further highlighting the potential application of deep learning technology in financial market decision-making.