Wine is an intriguing natural product involving biotransformation of grape constituents and continued chemical evolution during aging. Deep knowledge of the chemistry and sensory properties of wine has been developed, but gaps remain and new challenges emerge. Boundaries continue to be pushed with sensometabolomics used to identify important chemical contributors related to sensory acceptance and product quality. Rapid methods and timely access to results are also paramount for a digitally transformed industry, but masses of data are generated. Machine learning is quickly emerging as a versatile and powerful approach for handling multidimensional data sets and modeling intricate wine phenomena.
Sáenz-Navajas et al. (Thu,) studied this question.