Aging discrimination of French cheese types based on the optimization of an electronic nose using multivariate computational approaches combined with response surface method (RSM)

Abstract

This research was conducted to classify different types of cheeses using an electronic nose (e-nose). The experiments were accomplished in three storage periods such as on days 1, 7, and 14. To classify and analyze the output response of the sensors, artificial neural network (ANN), principal component analysis (PCA), linear discriminant analysis (LDA), support vector machine (SVM), partial least square (PLS), principal component regression (PCR) methods and response surface method (RSM) were used. Based on the results, on days 7 and 14 of storage, ANN classified all different cheese types with high accuracy. Additionally, using lodging plot, MQ3, TGS2610, and TGS2610 sensors had the widest use to discriminate among different cheese types. Moreover, the LDA method classified different cheese types with 93.52%, 96.7%, and 94.44% accuracy for the storage periods. The Nu-SVM function also attained the classification accuracy of 89.81% and 88.88% in validation and training datasets, respectively. Of the two PLS and PCR methods, PLS method achieved the highest accuracy for predicting the odor pattern among samples. Also, TGS880, MQ-3, and TGS2610 sensors were selected as optimized sensors for analysis by RSM. Consequently, e-nose coupled with multivariate methods and RSM showed satisfactory capability to recognize French cheese.