Classification of Wheat Flour: a Case Study in a Food Company

Abstract

Currently, the highly competitive environment requires food companies to optimize their processes. In this context, we suggest using artificial neural networks to optimize a food company's wheat flour classification process. The database, made available by the food company, presents 7666 observations. An algorithm based on the MLP (Multilayer Perception) architecture was implemented in the Python programming language. The Grid Search Cross-Validation technique was used to optimize the hyperparameters of the neural network. Experimental results showed that the MLP model presents an accuracy greater than 95% and a Kappa index of 0.949.