A recently developed AI algorithm has discovered over 300 previously unknown planets, which were revealed in the data gathered by the now-defunct exoplanet-hunting telescope.
Data gathered by a now-defunct telescope that hunted for exoplanets was analyzed by an artificial intelligence algorithm to discover over 300 previously unknown exoplanets. Astronomers have observed hundreds of thousands of stars using the Kepler Space Telescope, NASA’s first dedicated mission to search for exoplanets outside our solar system. Even after the telescope’s demise, its catalogue of potential planets continues to yield new discoveries. The data is analyzed by human experts for signs of extraterrestrial life. Now, ExoMiner mimics that procedure in order to scan the catalog faster and more efficiently.
During its operation, which ended in November 2018, the telescope looked for temporary decreases in brightness that might be caused by a planet crossing in front of the star’s disk from Kepler’s perspective. A NASA statement noted that not all of these dimmings are related to exoplanets, and scientists had to follow elaborate testing procedures to identify the real stuff from the false positives.
ExoMiner, is a so-called neural network, a type of artificial intelligence algorithm that can learn and improve its abilities when fed a sufficient amount of data. And Kepler generated plenty of data: In the less than 10 years of its service, the telescope discovered thousands of planet candidates, nearly 3,000 of which have since been confirmed. That is a vast majority of the overall 4,569 exoplanets currently known.
Over the course of its less than 10-year life, Kepler discovered thousands of planet candidates, of which more than 3,000 have since been confirmed. This represents a great deal of the current total of 4,669 exoplanets. This is due to the algorithm which is an example of an artificial intelligence algorithm known as a neural network, which is able to learn and improve as it is fed with enough data.
When looking at Kepler data, scientists look at each candidate exoplanet’s light curve and calculate the amount of the star that the planet seems to cover. As well as analyzing the apparent time it takes the would-be planet to cross the star’s disk. There are times when brightness changes observed cannot be attributed to an orbiting exoplanet. Researchers were able to add 301 previously unknown exoplanets to the Kepler planet catalog using the ExoMiner algorithm, which follows the same process but more efficiently.
“ExoMiner is highly accurate and in some ways more reliable than both existing machine classifiers and the human experts it’s meant to emulate because of the biases that come with human labeling. When ExoMiner says something is a planet, you can be sure it’s a planet.,” Hamed Valizadegan, ExoMiner project lead and machine learning manager with the Universities Space Research Association at the NASA Ames Research Center, said
Following the success of ExoMiner, scientists are hoping to use it for other exoplanet-hunting missions, such as NASA’s Transiting Exoplanet Survey Satellite (TESS) and the European Space Agency’s long-term PLATO mission.
In spite of the new confirmations, none of the newly discovered exoplanets are likely candidates for containing life since they reside outside of the habitable zones of their parent stars.
In a statement, NASA announced that the paper had been accepted for publication in the Astrophysical Journal; a draft of the paper can be read on the preprint site arXiv.org.