Revolutionizing Taste Testing: Machine-Learning Models Predict Beer’s Flavor and Ratings with Accuracy

How AI has the potential to revolutionize beer making

Researchers have developed 10 machine-learning models that can accurately predict the taste, smell, mouthfeel, and consumer rating of a beer. These models were created by combining chemical analysis with sensory features, allowing researchers to understand how factors such as flavor compounds and chemical properties influence overall taste.

To develop these models, researchers spent five years analyzing the chemical composition of 250 commercial beers. By measuring a beer’s chemical properties and flavor compounds, they were able to gain a deep understanding of how these factors affect taste. They then combined this information with assessments from a trained tasting panel and over 180,000 reviews from RateBeer, an online platform for beer enthusiasts.

Using this extensive data set, the researchers trained their models to accurately predict a beer’s taste, smell, mouthfeel, and consumer rating. This integration of chemical data with sensory features has the potential to transform the way food and drink products are developed and marketed in the future. With these advanced models, companies could save significant time and money that would have been spent on running trials to determine consumer preferences.

Leave a Reply