articles+ search results

In the present work, five models were evaluated in their ability to predict the kinetics of acid production in the lactic fermentation of sub-fishery products. The non-structured models evaluated were from the Gompertz, Baranyi-Roberts, Özilgen, Peleg and Vasquez-Murado models. Statistical evaluation of models included the Residual Sum of Square (SCE), Fisher's Test and Ross's Bias and Precision Indices. The models that presented lower values of SCE, absence of significant differences between them (p < 0.05) and better Bias and Precision Indices were the empirical models of Gompertz and the mechanistic model of Baranyi-Roberts, due to their ability to model symmetric and asymmetric curves in the first case, and their flexibility to different conditions in the second case. The Peleg empiric model and Vasquez-Murado logistic model failed to provide adequate fit for the type of lactic fermentation analyzed. The difference in the predictive capacity of the tested models was due to the fact that the fermentation was carried out with a starter of two strains with different rates of production of lactic acid, which generated a production curve with an asymmetry that the models of Peleg and Vasquez-Murado failed to fix properly. [ABSTRACT FROM AUTHOR]