Application of Stock Price Prediction during the "Double 11" Period Based on the LSTM Model

Studying stock price changes during special periods such as the "Double 11" shopping festival using the Internet E-commerce Industry Index can aid investors and e-commerce companies in making informed decisions. This paper focuses on stock price fluctuations during the "Double 11" period, using the "Internet E-commerce" Industry Index from January 20, 2016, to August 16, 2024, as the research object. The study employs ARIMA, LR, LGBM, and LSTM models to predict stock prices in the "Internet E-commerce" sector and uses RMSE, MAE, and R² as evaluation metrics to compare and assess model performance. The results confirm that the LSTM model demonstrates superior effectiveness and accuracy in predicting nonlinear stock price fluctuations, especially during special periods like "Double 11," compared to other models. Unlike traditional models, LSTM excels in capturing long-term dependencies and complex patterns in time series data. Its ability to adapt to nonlinear relationships and temporal variations makes it a more robust tool for stock price forecasting, particularly in volatile periods such as "Double 11".