Текст уведомления здесь

Russian machine learning helped CERN to save money allocated for searches for dark matter

Using machine learning, an international group of researchers with the involvement of Higher School of Economics and School of Data Analysis of «Yandex» have managed to lower substantially the costs of future experiment SHiP which will start in Europe in the 2020s. With its aid, the West scientists hope to detect the particles of dark matter which researchers failed to found for already many years. The respective article has been published in the Journal of Physics: Conference Series.
Добавить в закладки
Комментарии

The magnets for SHiP experiment will be produced in Russia. Their total weight will be about 2 thousand tons, while the created magnetic field will reach the value of 1.8 Tesla which is record-breaking number for «warm» magnets (not refrigerated by cryogenic devices).

In the new work, the Russian researchers have applied the method of machine learning to optimize the shape and position of magnets in «sifting» part of the device. By doing so, they have also managed to define their optimal mass. The algorithm of machine learning has generated a configuration which is by one quarter lighter than the basic one. This result has benefited in lowering detector costs by a quarter. The Russians have saved more than a million of dollars from CERN budget.

Добавить в закладки
Комментарии
Вам понравилась публикация?
Расскажите, что вы думаете, и мы подберем подходящие материалы