This research was conducted with the aim of analyzing the impact of selected Iranian financial markets - including the total stock market index, exchange rate, and gold coin price - on the food industry index, using a Long Short-Term Memory (LSTM) neural network model. The food industry, as one of the pillars of the country's economy, is strongly affected by macroeconomic and financial developments, and studying its dynamics can help identify inter-market structures and volatility transmission mechanisms.

For this purpose, time series data from March 2017 to February 2024 were used for modeling. The LSTM model, with its ability to identify nonlinear patterns and complex temporal dependencies, was employed to simultaneously predict both the level and volatility of the examined variables.

The results showed that the food industry index has the highest correlation with the total stock market index, while its relationship with the exchange rate and gold coin price - particularly at price levels - is weaker. At the volatility level, correlations between markets increase, indicating the importance of short-term shocks in market behavior.

Evaluation of the LSTM model based on statistical error indicators such as MSE and MAE showed that the model is capable of accurately reproducing the behavior of variables and has appropriate stability. Furthermore, the significant difference between price-level correlations and volatility correlations highlights the necessity of simultaneous analysis of these two dimensions in financial market studies.

The findings of this research provide a comprehensive picture of the dynamic structure of inter-market interactions and demonstrate the high efficiency of LSTM in predictive analysis of financial time series.