Well well well, finally some good news on this cursed rock. It seems beer is about to taste a whole lot better — at least, if the people behind the hops pay attention to the opinions of The Machines.
A research paper released today by Nature Communications, titled Predicting and improving complex beer flavour through machine learning, explores how machine learning models can predict and enhance the flavour of beer, based on its chemical properties.
In other words: computer make beer taste good.
The study involved analysing over 200 chemical characteristics of 250 Belgian beers across 22 different styles, along with sensory evaluations from a trained tasting panel, and data from over 180,000 consumer reviews on beer rating database RateBeer.
Kevin Verstrepen and colleagues used this large dataset to train 10 different machine learning models to predict which combinations of ingredients would elicit a better response from consumers. Normally these trials would be conducted with humans, which can be costly and time-consuming.
In other words: beer computer better than human consumer trials.
By analysing the best-performing models, specific and unexpected compounds were identified as key drivers of beer flavour and appreciation. Now here’s where it starts to matter to you — the addition of these compounds to commercial alcoholic and non-alcoholic beers resulted in improved consumer appreciation.
In other words: computer make beer taste good.
This research demonstrates how big data and machine learning can uncover complex relationships between food chemistry, flavour, and consumer perception, and lays the groundwork for further developments of novel, tailored foods with enhanced flavours
Cheers bro, I’ll drink to that.
