Student makes 1.5 million space discoveries after delving into old NASA archives

Matteo Paz was an elementary school student who fell in love with science and astronomy thanks to a series of talks at Caltech (California Institute of Technology).

Years later, as a more than dedicated high school student, fascinated by this unknown part of the universe and with artificial intelligence on his side, he took part in the Planet Finder Academy program to deepen his knowledge in astronomy and related computer science.

READ ALSO: He returned from a trip to Canada and ended up handcuffed: “I feared for my life”

It was his mentor, Davy Kirkpatrick, an experienced astronomer at IPAC (Infrared Processing and Analysis Center), who gave him access to the NEOWISE infrared telescope, which has been reliably scanning the sky for asteroids for over ten years and has also collected a gigantic amount of data on other objects.

The telescope recorded the thermal radiation of stars, quasars, and all types of pulsating and eclipsing lights (astronomers call them variable objects). The amount of information is so overwhelming, billions of entries, that deciphering it seemed nearly impossible.

So, together, but especially thanks to the student’s programming course, computer science theory, and even his advanced college-level math knowledge, they realized that this data, when organized, was perfect for artificial intelligence. In six weeks, he created his own machine learning algorithm to train the AI to search for telltale signs of variability in the telescope data, small but consistent changes in the brightness of objects over time that could reveal interesting cosmic phenomena.

READ ALSO: Warren Buffett warns Generation Z about making a fortune: “It’s not just hard work”

After a summer of work, the program identified and classified 1.5 million potentially new variable objects. A whole world of completely unexplored data from the Universe. A full catalog of the discoveries is now going to be published and made available to the entire scientific community.