Application of neural networks in the prediction of the circular economy level in agri-food chains

The objective of the work is to predict the level of circular economy in the agri-food chain through an empirical neural network approach. The research methodology includes the training of a neural network to predict the level of 128 circular economy in two agri-food chains. The novelty of this work lies in the possibility of defining in advance circular strategies based on the prediction of the level of circular economy. Historical data on the level of circular economy are compared with those predicted by neural networks. As a result, it is shown that if the weights of the circular economy level variables are not homogeneous, the procedure has a lower correlation value which, however, remains significant.