%0 Journal Article
%T 基于Transformer-LSTM模型的股票预测
Stock Prediction Based on Transformer-LSTM Model
%A 马昱晗
%A 孙德山
%A 代雨鑫
%J Advances in Applied Mathematics
%P 387-393
%@ 2324-8009
%D 2025
%I Hans Publishing
%R 10.12677/aam.2025.145268
%X 近年来，全球经济金融环境动荡不安，充满挑战与风险，在此情况下，投资者在金融交易方面的抉择更加复杂，股票交易市场也面临着更大的挑战，传统的计量经济学模型难以充分适应此类变化。本文运用深度学习中的长短期记忆(LSTM)神经网络作为基本模型对股票数据的收盘价进行预测，选用京粮控股(000505)、东北制药(000597)和中钢国际(000928)三组数据，并在LSTM神经网络中引入Transformer模型。经实验，该模型的预测精度有明显提升，说明Transformer-LSTM网络模型在股票预测领域具有一定的可靠性。
In recent years, the global economic and financial environment has been volatile and full of challenges and risks. In this situation, investors’ choices in financial trading have become more complex, and the stock trading market is also facing greater challenges. Traditional econometric models are difficult to fully adapt to such changes. This article uses the Long Short-Term Memory (LSTM) neural network in deep learning as the basic model to predict the closing price of stock data. Three sets of data from Jingliang Holdings (000505), Northeast Pharmaceutical (000597), and Zhonggang International (000928) are selected, and the Transformer model is introduced into the LSTM neural network. Through experiments, the prediction accuracy of the model has been significantly improved, indicating that the Transformer LSTM network model has certain reliability in the field of stock prediction.
%K 神经网络，
%K LSTM，
%K Transformer，
%K 股票预测
Neural Network
%K LSTM
%K Transformer
%K Stock Prediction
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=115577