Research on Stock Price Prediction Based on LSTM Neural Network and Markowitz Investment Model

This paper mainly focuses on the price prediction and trading strategies of domestic stocks, and constructs a stock price prediction model and investment strategy model by using entropy-weighted TOPSIS calculation, Bayesian optimization, LSTM neural network, Markowitz investment model and other methods. Firstly, the five stocks with the highest scores were selected by TOPSIS based on the entropy weight method, and then the stock price prediction of the above five stocks was carried out through the LSTM neural network model, and the Bayesian optimization method was used to optimize the model to ensure that the global optimal solution was obtained. Finally, MACD, Markowitz portfolio return theory and Sharpe ratio were used to solve the optimal portfolio of five stocks. This study provides an important theoretical basis and technical support for stock investors in price prediction and trading strategies.