Teen Discovers 1.5M Space Objects with AI, Wins $250K

A machine-learning model was created by a high school student to study archived space data collected by a retired NASA telescope. This model was trained to identify faint and hidden signals in the data—signals that had previously gone unnoticed.

The work was completed by Matteo Paz, a high school senior from Pasadena. During a summer research program at Caltech, his model was used to examine vast amounts of astronomical data. More than 1.5 million new cosmic objects—including stars, galaxies, and quasars—were discovered through the use of this algorithm.

The model worked by detecting small changes in light over time. These tiny shifts, often missed by traditional methods, were used to identify distant and dim space objects. The AI-based system allowed a faster and more accurate way to process the telescope's data.

The research was submitted to the Regeneron Science Talent Search, a prestigious science competition for high school students. Matteo's project was awarded first place, along with a $250,000 prize for his ground breaking discovery. His work also led to a first-author publication in a peer-reviewed journal, a rare achievement for a high school student.

Beyond astronomy, the model is now being considered for use in other fields. It could be applied to stock market analysis, climate research, and even air pollution monitoring, where subtle patterns in large datasets must be detected.

This achievement shows how machine learning, when used effectively, can unlock hidden information in old data—and how young minds can push science forward in unexpected ways.

📌Source : scitechdaily