ExoMiner, a new machine learning method helped discover 301 exoplanets

Advertisement
ExoMiner, a new machine learning method helped discover 301 exoplanets
ExoMiner has been trained using various tests and properties that humans use to identify and confirm new exoplanets.Unsplash
  • Scientists have discovered 301 new exoplanets with the help of a new deep neural network program.
  • ExoMiner was trained and fed with data from NASA’s Kepler spacecraft to identify new exoplanets.
  • It not only discovered new planets but also distinguished between the false ones.
A new deep neural network program called “ExoMiner” is helping scientists in discovering exoplanets. ExoMiner scoured through data from NASA’s Pleiades supercomputer to identify 301 unknown exoplanets. The machine learning method is also capable enough to distinguish between real exoplanets and imposters which are also called “false positives.”

NASA’s search for exoplanets was through its Kepler spacecraft which observed thousands of stars to identify possible exoplanets lurking around. It would look for temporary decreases in the brightness of the stars which would have possibly been caused by an exoplanet. Although NASA ended the mission in 2018, there is still a lot of data for scientists to read through and discover new exoplanets. This is humanly possible but it’s a tedious task, and this is where ExoMiner comes in.

ExoMiner has been trained using various tests and properties that humans use to identify and confirm new exoplanets. It also studies past confirmed exoplanets and those that are regarded as false positives. ExoMiner’s capabilities were recently proved when the machine learning algorithm validated 301 exoplanets from the Kepler Archive. These exoplanets were already detected by the Kepler Science Operations Center pipeline and even promoted to planet candidate status by the Kepler Science Office. And it was finally through ExoMiner that the final validation was confirmed.

Advertisement

Scientists discovered how ExoMiner is better at distinguishing between real exoplanets and the false positives. It’s not just picking out from the lot though as ExoMiner actually gives details on why the conclusion was made.

ExoMiner is considered more reliable than existing machine classifiers and human experts combined. Hamed Valizadegan, ExoMiner project lead and machine learning manager with the Universities Space Research Association at Ames, said that ExoMiner is now working with NASA’s Transiting Exoplanet Survey Satellite (TESS) to continue the search for more exoplanets.

SEE ALSO:

Advertisement
NASA’s planetary defence system: The mission that plans to defend earth from asteroids, comets
Jeff Bezos’s Blue Origin to take another tourist flight to space: Here’s all you need to know
{{}}