Comparison of Prediction Capability of Primary Models for Detection of Chicken Meat Spoilage

Abstract

The main objective of the present work is to compare the prediction capability of different primary models known as the modified Gompertz, modified logistic and Baranyi models to simulate the effect of temperature on aerobically-stored raw and marinated chicken meat spoilage using one-step modelling approach. For this purpose, total viable count (TVC) growth data were extracted from the published work for aerobically-stored raw and marinated chicken meat. The fitting capability of the global models was compared by taking into account root mean square error (RMSE) and adjusted coefficient of determination (adjusted-R2). Statistical indices, RMSE and adjusted-R2 values were found to be maximum 0.299 and minimum 0.970, respectively for each of the primary models and both of the chicken products. The prediction performance of the global models were evaluated with the rmax values that were independently published for aerobically-stored raw chicken meat, and RMSE values with lower than 5.11 × 10-2 revealed that one-step modelling approach can be reliably employed to predict TVC in aerobically-stored raw chicken meat.