presenting an intelligent stock price prediction model based on deep learning in tehran stock exchange market

Forecasting stock prices and returns is one of the most complicated and controversial topics in financial markets. stock market is constantly influenced by the state of the national economy, investors’ perceptions, and political events. furthermore, the price series is highly non-linear and unstable. ongoing research and updates in economic and stock market theories have gradually revealed the components necessary for predicting stock price indices, making accurate predictions possible. this research aims to develop an intelligent stock price prediction model based on deep learning for the tehran stock exchange market. this model incorporates dimensionality reduction techniques to manage the capital portfolio, thereby increasing returns and reducing investment risks. the data from 2020 to 2023 were sourced from the kodal system and were coded and analyzed using the risp method and the python programming language. a combination of lstm, pca, grp, and svd algorithms was used for the proposed model. a comparison of dimensionality reduction methods with artificial intelligence techniques shows that the pca dimensionality reduction method can enhance the performance of deep learning compared to other data dimensionality reduction methods.