LAUSR

Abstract An artificial (electronic) nose based on eight metal oxide semiconductor (MOS) sensors was developed and its capability to detect formalin, hydrogen peroxide and sodium hypochlorite in raw milk was investigated. Using chemometric approaches, feature vectors were extracted from the sensor array responses and used as the inputs into pattern recognition models. Principal component analysis (PCA) showed that PC1 and PC2 accounted for 97%, 87% and 83% of variance within data for formalin, hydrogen peroxide and sodium hypochlorite, respectively. Linear discriminant analysis (LDA) revealed the relatively low classification accuracy as 79.16%, 70.83%, and 66.66% for formalin, hydrogen peroxide and sodium hypochlorite, respectively. Finally, the support vector machine showed accuracy values of 94.64%, 92.85% and 87.75% for formalin, hydrogen peroxide and sodium hypochlorite, respectively. The results demonstrated that an artificial nose, combined with pattern recognition methods, could be used for detection of these adulterants in raw milk.