When galaxy clash with each other it ’s a very slow operation , which can take billions of years to complete . It is crucial for the evolution of galaxies but what we know come from look at different objects . We have to calculate at different snap   to seek to work up a full pictorial matter of what ’s going on .

A paper of late put out inAstronomy & Astrophysicsused artificial intelligence ( AI ) to study mergers from 12 billion year ago until today . A sum of 200,000 mergers , a truly record - break number , were analyze and the researchers found that while merging galaxies do n’t have particularly high headliner - shaping rates , they do lead to increase starburst form .

Starbursts are localized episodes of intense star formation and in the squad ’s sampling , there were twice as many starbursts in merging wandflower compared to in single galaxies .   The researchers stress how crucial the entry of neural connection algorithm has been to achieve their final result . This   tick off the first fourth dimension the approach has been used to study galaxy fusion .

" This is a milepost in the sense that AI will diddle an more and more large persona in our field , " squad drawing card Dr Lingyu Wang , from SRON Netherlands Institute for Space Research , enunciate in astatement . " But we have to keep in mind that the baron of AI is limited to how it is trained . If we feed it a blemished definition of a galaxy merger , then it wo n’t do its Book of Job right . "

The limitations of AI have been a significant issue in implement these attack . For many years , humans were much better at spotting the hallmark of galax mergers   than automated approaches . Now , AI seems to have caught up .

To treat with the huge database , the squad had do make a machine - learn algorithm that taught itself what merging galaxies looked like . It was good enough to deliver a very round-eyed but effective binary classification of merger / non - merger .

" The reward of artificial intelligence is that it improves the reproducibility of our written report because the algorithm is consistent in its definitions of a amalgamation . Also , it ’s good homework for upcoming surveys that will image billions of galax , "   first writer William Pearson explain . " Then you needs require AI . Even citizen science undertaking such asGalaxy Zoocannot deal with those numbers . "