Prediction of secondary components in sugarcane brandy by application of artificial neural network

This study aimed to use artificial neural networks to predict the secondary components of sugarcane brandy. We got data on the characteristics of sugarcane brandy from the literature. We divided this data into input data and output data. Secondary components in sugarcane brandy were the output data. The architecture used for artificial neural networking was the multilayer feed-forward network, which features a hidden layer. These hidden neurons have the role of intervening between the input and output layers of the network. We separated the data into 70% for training and 30% for tests. The artificial neural network transfer function was Relu, with Adam as the training algorithm for weight change with a constant learning rate. The number of neurons in the hidden layer was determined using the mean square error. Twenty neurons in the hidden layer were analyzed and considered appropriate, since in the artificial neural network with a greater number of neurons, we observed no significant variation in the reduction of error.