In a method known as reflectometry, scientists can characterize or detect objects by studying the reflection of waves on surfaces. There are several subsets of reflectometry, including neutron reflectometry, a technique used to measure the structure of thin films on materials.
Doctoral student Amanda Colunga helped develop a graphical user interface for neutron reflectometry data during an internship at Oak Ridge National Laboratory.
Analyzing neutron reflectometry measurements can be difficult, especially if the scientist isn’t familiar with how the data was collected or the material being examined.
Through the GEM Fellow Internship Program at Oak Ridge National Laboratory (ORNL), doctoral student Amanda Colunga took part in efforts to automate this data analysis, making it easier for scientists to interpret and interact with reflectometry measurements.
ORNL partners with the National GEM Consortium to host GEM fellows for summer research experiences. The National GEM Consortium is a network of leading corporations, research institutions and universities that enables qualified students from underrepresented communities to pursue graduate education in science, technology, engineering and mathematics (STEM) fields.
Under the guidance of Richard Archibald, Ph.D., in ORNL’s Computational and Applied Mathematics Group, Colunga and two other interns sought to build a graphical user interface (GUI) that will aid in the analysis of reflectometry data.
“This tool will generate the most likely models for the data given, allow scientists to fix parameter values and add boundaries, and optimize those parameters and give the confidence intervals for which the true values likely are,” Colunga explained.
To sort through the data, the team turned to machine learning, a subset of artificial intelligence. Machine learning involves the development of computer algorithms that give systems the ability to improve or “learn” through experience, without being explicitly programmed. For Colunga’s project, the team employed the K-nearest neighbors algorithm, a machine learning method that classifies new data points based on data that has been provided and labeled.
Through her internship, Colunga learned the ins and outs of the Python™ programming language, which was used to build the GUI. She had some prior experience using MATLAB™, another programming language, and she was able to apply that coding knowledge to Python.
Colunga regularly met with the fellow interns on her team to discuss the project and ways to make improvements. For Colunga, the collaborative nature of the research was a highlight of the internship.
“I am a big proponent for working in groups because we can all learn new things from each other,” Colunga said. “At ORNL, I collaborated with an undergraduate student and a high school student, and I learned as much from them as they learned from me. Having the opportunity to collaborate on this project really made my experience at ORNL.”
Overall, Colunga said her time at ORNL was a great experience that exposed her to different areas of science and research.
“I never thought that I would have had the opportunity to conduct research at a national lab, so I am grateful for the opportunity that GEM and ORNL gave me,” Colunga said. “I was amazed at the kind of science that was being done at ORNL and the collaborations between the departments.”
Colunga is pursuing a doctoral degree in applied mathematics at North Carolina State University. Her research focuses on using modeling and differential equations to gain insights on the human cardiovascular system.
Her advice to future ORNL interns? Get outside your comfort zone and don’t be afraid to ask for help.
“We are all learning, and ORNL has people who are more than happy to assist,” she said.
The GEM Fellow Internship Program at ORNL is administered by the Oak Ridge Institute for Science and Education (ORISE) for the U.S. Department of Energy.