Artificial intelligence—imagine all of the ways this technology could help with the advancement of mankind! And in lieu of that, imagine how it could be used to track the prices of fine wine on the secondary buying market.
Innovation often follows the money, so it probably shouldn’t be surprising that researchers from the University College London, who have previously applied artificial intelligence algorithms to the medical and financial worlds, decided to dabble in the high-stakes game of fine wine trading. In a study published this week in the Journal of Wine Economics, the team explained how their newest machine learning methods have proven more accurate than traditional handicapping in determining how the prices of the world’s top wines will fluctuate.
“We've shown that price prediction algorithms akin to those routinely used by other markets can be applied to wines,” Tristan Fletcher, co-author of the study, was quoted as saying by Phys.org. Fletcher has a lot riding on this work: Not only is he an academic at UCL, he’s also founder of Invinio—a quantitative wine asset management firm.
For the study, researchers tested out two forms of machine learning on 100 of the most sought-after wines included on the Liv-ex 100 wine index. Both methods proved fruitful, albeit in different ways: One increased overall average accuracy by 15 percent, whereas the other increased accuracy by 98 percent but only on about half the wines.
“We hope our findings give the industry confidence to start adopting machine learning methods as a tool for investment decisions,” said primary author Michelle Yeo. All right, sure, but when can we drink the wines? That’s what my non-artificial intelligence thinks we should do.