Abstract: Principal component analysis (PCA) and Fisher linear discriminant analysis were both applied for the pattern recognition and classification of Zengcheng mew rice from Guangdong province as a geographical indication (GI) product. A classification model between GI and non-GI products was established based on inorganic elements and organic compounds including amino acids, fat, starch and protein by Fisher linear discriminant analysis. The results showed that Fisher linear discriminant analysis was a simple and fast method for the classification and identification of geographical indication rice and could allow rapid establishment of a classification model for identification and reorganization between GI and non-GI products based on relatively complicated data.

Key words: national geographical indication product, rice, pattern recognition, Fisher linear discrimination