Modeling of tempering conditions depending on chocolate components using artificial neural networks

The tempering quality of the chocolate has a significant impact on the expected appearance and textural properties of the final product. Consequently, it is a key parameter for consumer acceptance. Determining the tempering temperature and quality of chocolate requires specialized equipment, which creates personnel and time demands. The melting profile indicates the melting temperatures of different crystal forms in chocolate, and this information guides the determination of tempering temperatures. Chocolate manufacturers often prefer to produce samples to understand the quality and the entire process in new formulation studies. Chocolate is a multi-component system that includes cocoa butter and, depending on the type, milk fat. An artificial neural network that can predict the tempering temperature of chocolate will facilitate this process. Artificial Neural Network (ANN) modeling was applied to predict the tempering temperature based solely on chocolate ingredients. The ANN models were able to predict the properties with good accuracy (R² > 0.96) and low mean square error (MSE). This study demonstrated its effectiveness as a tool for predicting the tempering temperature of chocolate based on its ingredients. Direct predictions can be made without creating a melting profile. This allows for rapid decision-making, particularly in R&D processes.