The project of American schoolboy Matteo Paz, who used machine learning to sift through NASA's vast archive of observations, has grown into a full-fledged astronomical catalog. VarWISE has already collected more than 1.9 million space objects with variable brightness. Moreover, the authors consider more than 82% of them to be new discoveries.
The project of American schoolboy Matteo Paz, who used machine learning to sift through NASA's vast archive of observations, has grown into a full-fledged astronomical catalog. VarWISE has already collected more than 1.9 million space objects with variable brightness. Moreover, the authors consider more than 82% of them to be new discoveries.
This is reported by Space Daily.
Varying brightness is when a celestial object gets brighter or dimmer over time. This can happen, for example, due to stellar pulsations, flares, or when one object periodically obscures another. The work on VarWISE was published in 2026 in The Astrophysical Journal Supplement Series. The catalog is based on data from the NEOWISE telescope, which has been scanning the entire sky in infrared light for more than a decade.
The algorithms that power the catalog try to understand how each object's brightness changes, assign it to a specific type, and for objects with regular changes, even determine their cycle. So the computer takes on the work that a person would be almost impossible to do manually: it looks through a huge stream of observations and looks for places where something in space regularly pulses, flashes, or, conversely, dims.
That's where Paz's project began. While still a schoolboy, he wrote an algorithm that analyzed about 200 billion individual NEOWISE records. In 2025, the program selected about 1.5 million potentially new objects, and for his work, Paz received the top prize of $250,000 at the Regeneron Science Talent Search. In addition, after the school project, he became an employee of IPAC - a center that, among other things, works with data from NASA space missions.
Back then, it was primarily a long list of candidates that still needed to be vetted. Now, the project has evolved into a separate scientific catalog, available through the NASA/IPAC Infrared Science Archive, so other astronomers can use it for their own research.
Previously, dev.ua wrote that Veronika Khomukha, a 15-year-old schoolgirl from Kryvyi Rih, created algorithms that help amateur astronomers search for variable stars.
