Premium
Classification of Rice Amylose Content by Discriminant Analysis of Physicochemical Properties
Author(s) -
Suwannaporn Prisana,
Pitiphunpong Sawidtree,
Champangern Sirirat
Publication year - 2007
Publication title -
starch ‐ stärke
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.62
H-Index - 82
eISSN - 1521-379X
pISSN - 0038-9056
DOI - 10.1002/star.200600565
Subject(s) - amylose , linear discriminant analysis , food science , mathematics , texture (cosmology) , quality assessment , chemistry , pattern recognition (psychology) , artificial intelligence , computer science , statistics , evaluation methods , starch , engineering , reliability engineering , image (mathematics)
Amylose content is commonly used to predict the texture of cooked rice. However, rice classification using only amylose content can misclassify actual sensory quality. Quality assessment of cooked rice can be determined by a combined evaluation of physical, chemical and sensory properties, but this process is costly and requires use of trained panelists. In this study, an alternate method for predicting rice amylose content was investigated. Paste viscosity and cooked rice texture of nine rice varieties were measured that had either low, medium, or high amylose content. Data were analyzed using analysis of variance, Pearson correlation, and discriminant analysis. Results showed that protein content, which played a major role in cooked rice texture, did not correlate with amylose content. Discriminant analysis results showed that only pasting attributes could discriminate the amylose groups but not texture. Pasting property, peak viscosity, and trough were the best discriminators with an eigenvalue of 32.9. These discriminators correctly predicted 100% of medium‐ and high‐amylose rice, but only correctly predicted 66.7% of low‐amylose rice with 88.9% of cross‐validated grouped cases correctly classified. The use of rice texture and pasting properties in combination with multivariate analysis techniques has potential as an easier and less costly method for the accurate classification of rice based on sensory quality.