Application of LSTM and Attention Mechanism for Stock Price Prediction and Analysis

The relationship between stock prices and economic development is well-established, and the study of stock price prediction methods is crucial for gaining insights into the economy.This research aims to enhance the accuracy of stock price prediction by leveraging a combination of convolutional neural network (CNN), attention mechanisms, and long short-term memory (LSTM).Specifically, the stock data of Banco Bilbao Vizcaya Argentina (BBVA) was selected as the research object, and the stock price prediction results after incorporating efficient channel attention (ECA), channel attention module (CAM), spatial attention module (SAM), and convolutional block attention module (CBAM) in CNN + LSTM were analyzed.The findings demonstrate that the incorporation of attention mechanisms in CNN + LSTM has a positive impact on stock price prediction accuracy.Notably, the model that integrates CBAM attention mechanism in CNN + LSTM yields the best prediction results, with MAE, MSE, RMSE, and R 2 metrics achieving 58.68 × 10 -4 , 7.94 × 10 -5 , 8.91 × 10 -3 , and 0.9673, respectively.These results have implications for improving the accuracy of stock price prediction and provide valuable insights for future research in this area.