An image of Sagittarius A *, the black hole in the center of GaAssy Way Galaxy
and
Our Galaxy’s Center may have a unique cosmic spinning top – a black hole as if wasting is almost as fast as possible.
Michael Janssen To the Radboud university in the Netherlands and his companions studied at Black hole in the center of the Milky WaySAGITTARIUS A *, using the data assembled by a network of collective observators known as the Telescope event (EHT). To take care of the complicated data, they have made artificial intelligence.
First, they used the well-known mathematical models to glorify about a million black holes – which itself a computation fee needed at the time of supercomputer. Then they use these simulations to train a class AI called a Neural networkthat allows this to determine a black hole character-based attitudes to observation. In the end, they fed AI the data about Sagittarius A * collected by Eht throughout 2017.
AI indicated that Sagittarius A * wasted between 80 and 90 percent of the highest possible speed. It also alerted researchers with no one of their magnetic models that ended well our black hole – must be more mathematical work. Jansen said that previous studies prevented the order of the Sagittarius A *, as it was as easy to spin the Magnetic fields surrounds itBut this new method focuses on them the more accurate.
Dimitrios Psaltis At the Georgia Institute of Technology in Atlanta says he found some of these opaque opponents. The previous analysis could not even provide clarity if the black hole in the hole can determine it accurately from Eht data, he said.
Some previous jobs show that Sagittarius A * could make it easier, as Yosuke Mizuno In Shanghai Jiao Tong University in China. But he noticed that the computer models used in the new study have a place for prosperity. “Our theoretical model is still not perfect,” he said.
But both Mizuno and Psaltis say AI has become a significant part of it how we know exotic cosmic things like black holes. “We are in a situation where we have a lot of data and we have a lot of models, and we need the modern ways of uniting both,” says Psaltis. “This is where machine learning makes a big, big difference.”
At the same time, it brings his own challenges, because AI employment should double checked and subsequent evidence confirmed against possible repentances.
Janssen and his team have already made many such checks, such as AI testing with specified simulation data. Many attempts occur as they analyze data from other years of EHT operations and, finally, data from new observatures, he said.
Topics:
