A research team from the University of Warwick has confirmed the existence of 118 exoplanets. The project uses artificial intelligence to analyze large volumes of astronomical data. The team members applied RAVEN, which is a machine learning algorithm designed to filter spatial signals.
This computer system processed data previously collected by NASA’s TESS satellite, which monitors the brightness of stars to detect planetary transits. A planetary transit is a temporary decrease in stellar brightness, caused by the passage of a celestial body in front of the star.
Therefore, identifying these fluctuations requires high calculation accuracy. The RAVEN algorithm separated the actual fluctuations in light from the false cosmic anomalies caused by binary stars, which leads to accelerating the validation of planets in databases.
The system analyzed thousands of light signals, with the algorithm identifying planet-specific patterns. Thus, the time allocated to the manual analysis of light curves has been considerably reduced. The algorithmic method demonstrates a higher success rate than traditional statistical techniques of astronomical validation.
Sources:
- https://warwick.ac.uk/
- https://unn.ua/en/news/artificial-intelligence-found-dozens-of-hidden-planets-in-nasa-data
Photo by NASA Hubble Space Telescope