Identifying the Primary Odor Perception Descriptors by Multi-Output Linear Regression Models

Semantic odor perception descriptors, such as “sweet”, are widely used for product quality assessment in food, beverage, and fragrance industries to profile the perceptions. The current literature focuses on developing many possible descriptors. A large number of descriptors poses challenges sensory assessment. In this paper, we propose task narrowing down To end, contrive a novel selection mechanism based machine learning identify primary perceptual (POPDs). ratings non-primary (NPOPDs) could be predicted precisely from those POPDs. Therefore, NPOPDs redundant disregarded vocabulary. experimental results indicate that dozens redundant. It is also observed sparsity data has negative correlation coefficient with model performance, while Pearson between perceptions plays an active role. Reducing vocabulary size simplify auxiliary understand human space.