Predictive Analysis of the Agricultural Price Index using Artificial Intelligence

The research addresses the problem of price volatility in the Peruvian agricultural sector, which affects economic stability and food security.The main objective is to develop a predictive model using LSTM neural networks to anticipate price variations and provide useful tools for decision making in the sector.The methodology included the normalization of historical data obtained from the Ministry of Agrarian Development and Irrigation, followed by the implementation and training of an LSTM model, evaluated using the root mean square error (RMSE).The results showed a training RMSE of 1.9904 and a significantly higher test RMSE of 48.0675, indicating a possible over-fitting of the model.The future price index predictions were 144.0155 and 124.3000, suggesting downward trends not fully aligned with historical data.In conclusion, although the LSTM model was able to capture patterns in the training data, its generalizability was limited, highlighting the need to improve the model and data quality to obtain more accurate and robust predictions in agriculture.