Optimizing the stock portfolio using the technical mean-variance method and random forest

The purpose of this research is to optimize the stock portfolio using the technical mean-variance method and random forest (RF). The statistical population of this research includes all companies active in the Tehran Stock Exchange during the years2011 to 2023. In this research, 40 active stocks in the stock market that had continuous and complete data were selected. In order to check the optimization of the stock portfolio, the stock price information of 5 companies admitted to the stock exchange was used. In order to optimize the mean-variance of the stock portfolio based on the analysis of technical signals, the relative strength index (RSI) of the leading indicators was used and for modeling two methods of mean-variance technical Markowitz analysis and algorithm Random forest is used. To evaluate the performance of the proposed model, mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) have been used. The obtained results show that the values ​​of the proposed indices for the technical mean-variance model network have lower values ​​compared to the RF algorithm. And this issue shows the higher accuracy of this recurrent neural network than the random forest algorithm in modeling and predicting the optimization values ​​of the stock portfolio.