APLICAÇÃO DE MACHINE LEARNING NA AVALIAÇÃO DA INFLUÊNCIA DO PROPIONATO DE CÁLCIO EM PÃES
APLICAÇÃO DE MACHINE LEARNING NA AVALIAÇÃO DA INFLUÊNCIA DO PROPIONATO DE CÁLCIO EM PÃES
-
DOI: https://doi.org/10.22533/at.ed.3952424104
-
Palavras-chave: Alimentos, aprendizado de máquinas, shelf life, sensorial, JAR
-
Keywords: Food, machine learning, shelf life, sensory, JAR
-
Abstract: This study explores the influence of calcium propionate on the sensory and physical properties of bread, as well as the use of machine learning models to determine this influence. Two bread formulations, one with and one without calcium propionate, were prepared and analyzed over a 21-day period. Texture parameters such as hardness, elasticity, and gumminess were measured, and sensory analysis was conducted using the Just-About-Right (JAR) scale. The data were processed using Random Forest, Logistic Regression, and SVM models, with SHAP values applied to interpret the impact of different variables on the predictions. The results show that calcium propionate prolongs the shelf life of bread but significantly affects its texture. The Random Forest model achieved the best balance between precision and recall, demonstrating its effectiveness in determining the presence of preservatives. These findings suggest that combining sensory analysis with machine learning techniques can be a powerful approach in food quality assessment
- Márcia Arocha Gularte
- LAYLA DAME MACEDO
- Bianca Pio Ávila
- Roberta Bascke
- Aline M. Pereira