MTUCI Neural Network Learns to Predict Automakers’ Stock Prices
Researchers at the Moscow Technical University of Communications and Informatics (MTUCI) have developed an AI model that forecasts automakers’ stock prices, even during sharp market swings.
The system is built around a hybrid machine-learning model that combines two types of recurrent neural networks, GRU and LSTM, in sequence to analyze complex financial time series. The algorithm processes historical stock-market data automatically and generates price forecasts.
The system can be integrated into existing financial analysis tools, banking platforms, and IT systems used for securities-market analytics.
“Testing has confirmed the model’s practical value. It could be used by investors analyzing companies in the real economy, as well as in higher education to train specialists in financial modeling,” the university said.
The developers plan to incorporate a wider range of economic factors and adapt the model for related sectors of high-tech manufacturing.
