An electronic nose system for assessing horse mackerel freshness

An electronic nose is developed and applied to freshness test of the horse mackerels. 8 metal oxide gas sensors are used in the electronic nose. The data obtained from the electronic nose are processed with baseline manipulation, normalization, feature extraction, feature subset selection and classification stages respectively. Only one feature is extracted from each sensor. So every odor has 8 features. The classification of freshness is implemented with k-nearest neighbor, artificial neural network and decision tree methods and the success rates of three methods are compared with each other. It is observed that decision tree method is the best of these tree methods for freshness classification of the horse mackerels.