Network Structure to Estimate Prices of Basic Products: Dairy

Price prediction is one of the key issues in sector analysis. However, both the level of production and the prices of the agricultural sectors are highly variable characteristics with strong dependencies on eventualities, since they are subject to climatic and political shocks, making complex the modeling of their behavior and, consequently, the task of predicting or forecasting their future evolution. The present study aims to adjust a neural network model to make predictions applied to the Colombian dairy sector and to compare the results with the predictions provided by a multivariate time series model.