Scientists have to tackle all kinds of different problems. Not all of them look important for everyone, but new discoveries are important nevertheless. One of such seemingly insignificant problems is fluctuations in market of fine wines. Now scientists at University College London have developed a more accurate way of predicting prices of fine wines. Their approach is based on using a novel artificial intelligence system.
And this problem is actually quite important to solve. It will help fine wine investors to make more informed decisions about putting money into the market of fine wines as well as will make such investments more popular between people who never tried it. Furthermore, scientists expect to use this approach in predicting prices of other assets as well, such as classic cars.
Dr Tristan Fletcher, one of authors of the study, said: “People have been investing in wine for hundreds of years and it’s only very recently that the way they are doing it has changed. Wine investment is becoming more accessible and is a continually growing market, primarily brokered in London: the world-centre of the wine trade. We’ve shown that price prediction algorithms akin to those routinely used by other markets can be applied to wines.”
Price prediction algorithms are used already to predict fluctuation in financial markets. However, they did not work for such specific market as fine wines. Therefore, scientists came up with more complex machine learning methods, which outperformed other simpler processes commonly used for financial predictions. New approach was applied to 100 of the most sought-after fine wines from the Liv-ex 100 wine index and gave surprisingly good results. It managed to predict prices with greater accuracy than any other machine used before, because it can learn which information is important amongst the data.
That is why it is called an artificial intelligence approach, because machine can learn by itself, which involves developing algorithms that automatically learn from new data without any human intervention. Machine searches through data, extracts and uses useful information to predict the values of fine wines. Scientists have been developing systems like this for quite a while and such methods have been already used in such industries as medical and financial and now it will be used to predict prices in a market of fine wines. However, as usual, it was not an easy task to create such artificial intelligence that would be able to choose important information and use it by itself without human intervention.
Team of researchers tested two forms of machine learning including ‘Gaussian process regression’ and the more complex ‘multi-task feature learning’. Both of these machine learning techniques are very advanced and are able to extract the most relevant information from a variety of sources. Meanwhile, more standard machine learning processes typically assume every data point is of interest, spurious or otherwise. The two forms of machine learning that scientists chose to use have different advantages.
‘Gaussian process regression’ can be applied to a wide variety of wines. In fact, it can be applied to all the wines in the Liv-ex 100 and achieves improvement of accuracy by 15% over standard predictions. Meanwhile the ‘multi-task feature learning’ learning model can only be applied to half of the wines analysed as it required a stronger relationship between prices from one day to the next, but achieves much more significant 98% increase in accuracy over the standard benchmarks.
Scientists hope that this new method for predicting price fluctuations of fine wines will give entire marker more confidence and will encourage investors to participate more in purchasing these wines. It also should attract new people who were not confident about their expertise in the field. Of course, some work remains to be done, as team will continue to improve the algorithm both in terms of performance and accuracy. And later they will try to apply it in the world of classic cars that are also very attractive for people searching for safe and profitable investments.