Application of LSTM and portfolio optimization in Chinese stock market

The purpose of this paper is to examine the application LSTM and mean variance portfolio optimization in Chinese stock market. 20 stocks are selected from CSI 300 components, we collect their High, Low, Open, Adjust Close trade volume June 16th 2020 2022. Then use model forecast price. results used construct 2 portfolios. One maximize Sharpe ratio, other minimize variance. From April 6th 2022, Maximize Ratio outperformed index, Minimize Variance Portfolio did not beat market but return was very close CSI. Therefore, combination Mean Optimization theory effective