AI will now help find galaxies in deep space
Artificial Intelligence has one more skill added to its ‘brain’. Now it can speed up the process of finding galaxies not known to science.
Artificial Intelligence can play games, drive a car, and fly a drone. Now, we can add one more skill to the list: finding galaxies.
Researchers at Lancaster University are developing “Deep-CEE” (Deep Learning for Galaxy Cluster Extraction and Evaluation). It is a deep learning technique that is expected to speed up the process of finding galaxies.
Why is it difficult to find galaxies?
Galaxies are found in small groups like our Milky Way and the neighbouring Andromeda galaxy. Most of the galaxies in the universe live in low-density environments called "the field". Galaxies rarely exist in clusters, as clusters are the most extreme environments that galaxies can live in. Studying these galaxies can help us know more about dark matter and dark energy.
How will AI find Galaxies?
It can be as simple as looking at images.
Matthew Chan, a Ph.D. student at Lancaster University, who has been working on this project, used images of galaxies to train the AI. The AI model has been trained to 'look' at the 'colour' images of existing galaxies.
‘Deep-CEE' is based on a neural network. It mimics the learning pattern of the human brain. This process similar to the human brain activates specific neurons when visualizing distinctive patterns and colours.
George Abell, the pioneer of galaxy cluster-finding, used to find galaxies using the eye, a magnifying glass, and photographic plates. He analysed around 2,000 photographic plates and found what we now know as the ‘Abell catalogue’ of galaxy clusters.
The approach that is used by AI is similar to how Abell used to find galaxies.
“We have successfully applied Deep-CEE to the Sloan Digital Sky Survey,” said Chan. “Ultimately, we will run our model on revolutionary surveys such as the Large Synoptic Survey Telescope (LSST) that will probe wider and deeper into regions of the Universe never before explored."
So, are telescopes useless now?
No. We have built state-of-the-art telescopes that enable astronomers to look deeper than ever before. But, AI is expected to automate the process bringing down the level of human interaction. There is no longer a need for human scientists to gaze at the sky for years to find a galaxy.
This will also be helpful in analysing data for us in the future. The new LSST sky survey, which will likely start in 2021, is expected to generate 15 TB of data every night. This amount of data can be analysed by AI.
“Data mining techniques such as deep learning will help us to analyse the enormous outputs of modern telescopes,” says Dr. John Stott (Chan’s Ph.D. supervisor). “We expect our method to find thousands of clusters never seen before by science.”