Rapid quality assessment of bread using developed multivariate models: A simple predictive modeling approach

Abstract

AbstractThis study was focused on the assessment of relationships among the properties of wheat and their resultant flour, dough and final bread. For this purpose, multivariate linear regression in the form of the step-wise algorithm was applied to evaluate the relation among the flour characteristics of wheat with quality of dough and the final breads (Barbari and Lavash). The results showed that variety of wheat (Orum, Pishgam, and Zareh) could not affect the moisture content and quantity of the flour residue; however, considerable variation was observed on protein content and Zeleny number. The multivariate regression analysis built appropriate models to predict the hardness of the Barbari bread (R2 = 0.98) and specific volume of the Lavash bread (R2 = 0.98). Overall, the results indicated that the regression models in the form of step-wise might be useful as a non-destructive technique for assessing quality of bread.