Abstract: Based on the sensory evaluation combined with self-organizing maps (SOM) model, the preference of sausage products was predicted. The texture parameters and color data of 99 kinds of sausage products were collected as the research objects, and the linear regression analysis and correlation evaluation were carried out on the sensory evaluation results. The principal component analysis method was used to eliminate the redundant data. A SOM model with a competition layer of 6 and an output layer of 36 was established. The results showed that the accuracy rate was 100% by extracting and classifying the eigenvalues of sausage samples. At this time, the RMSE of the prediction set was 0.1184, which implied that the model showed good generalization ability. This study aims to establish an accurate and efficient method for predicting food preference, providing a data reference for food product development and market preferences.

Key words: Self-organizing maps, Sausage, Preference, Prediction