Charles Zhao: extracting complex data from high-powered telescopes
Charles Zhao is majoring in computer science and is slated to earn two certificates: Statistics and Machine Learning from the Center for Statistics and Machine Learning (CSML) and Engineering Physics.
In his CSML independent project, which he completed in his junior year, Zhao applied machine learning techniques to extract information from images captured by large, powerful telescopes. Specifically, Zhao looked at galaxies’ central nuclei, which consist of a black hole and other astronomical phenomena and features. In this nucleus, there are energized regions of gas that produce radiation. These regions of gas are called extended emission-line regions (EELRs). His project consisted of extracting EELRs from noisy, cluttered images captured by telescopes.
Zhao used a Gaussian Process regression to generate samples of likely EELR spectra, and then computed a likelihood-weighted model average of each EELR’s morphology and spectrum to produce the final estimator. Zhao, whose advisor is Peter Melchior, assistant professor of astrophysical sciences and CSML, also used Melchior’s algorithmic method, SCARLET, which allows for the morphologies and spectra of individual astronomical sources to be extracted from large data sets even when they are partially or completely overlapping.
Zhao tackled this project because a method was needed for picking out EELRs in images where galaxies’ own radiation often dominated that of the EELRs.
“I wanted to do this because I’ve always been very interested in physics and astronomy,” said Zhao.
For his undergraduate research and his future endeavors, Zhao has found the CSML certificate to be very helpful. He plans on entering the machine learning field after graduation as an autonomy software engineer for the self-driving company, Nuro, where he interned last summer.
“I’ve been interested in self-driving cars for a long time, since high school. I think self-driving cars have a very clear real-world impact and they're also just a lot of fun. I like working on real world applications with interesting problems,” said Zhao.
Other projects he has worked on include an AI-powered motivational mood tracker and applying machine learning in exoplanet detection.
Besides his experience interning at Nuro, Zhao worked as a quantitative research analyst intern at Stevens Capital Management, a computer science and mathematics division intern at Oak Ridge National Laboratory, and software engineer at PrepFactory.
Zhao is the data science lead for NJ Student Climate Advocates. In this organization, he leads a team of students in performing data science projects, and helps develop policy proposals and research.
Zhao enjoys rock climbing, playing the piano, and dabbling with 3D animation.