The influence of data preprocessing on the correlation between instrument and sensory evaluation of rice

• 1、School of Food Science and Technology,Jiangnan University,Wuxi 214122

Abstract：21 kinds of commercial rice were selected, the texture of rice was determined by TPA mode, the gelatinization characteristic of rice flour was determined by rapid viscosity analyzer, the sensory evaluation of rice taste quality was conducted according to the national standard. The correlation analysis and regression model were established for texture results, gelatinization characteristic results and sensory evaluation. Compared the results of data scaling and log conversion of the original data and the results of original data, the correlation analysis and regression equation of the transformed original data have lower standard estimation errors and the same significance.By transforming the data of different instruments, the difference of data order and method between instruments and measurement methods is eliminated, which provides the possibility for the universality of sensory evaluation regression equation. Correlation analysis and regression equation show that there was a significant negative correlation between the peak viscosity, regeneration value and sensory evaluation flour amd a significant positive correlation between cohesiveness and sensory evaluation.

Keywords： Rice Physical and chemical properties Sensory properties Data processing Multiple linear regression