Special to the E-T

Growing up in southern California, Katherine Leaveck dreamed of a career in science.

“It’s hard to say when I first became interested in astronomy because I have always loved science, especially space,” said Leaveck, 26. “I know in the fourth grade I told my dad I was going to go to college to study science, so space has been my passion since before I can remember.”

The Leaveck family moved to Dublin, Texas, when Katherine was in junior high. After graduating high school, she chose to pursue her dreams and her college career at Tarleton State University.

For a relatively small school, Tarleton has made huge strides in the area of research and development. One project, in particular, may one day change the way astronomers study stars.

Leaveck is working to develop a program that will sift through large amounts of data and allow scientists to more easily and efficiently study groups of stars.

Recent technological advances have made it possible for astronomers to accumulate large amounts of data on binary star systems.

The term “binary stars” refers to a system in which two stars revolve around a common center of mass. Astronomers study binary stars to find the masses of individual stars and study how stars evolve.

While manually sorting through this information used to be a viable option, the drastic increase in the amount of data collected called for an automated classifying system.

Leaveck, a full-time student, spends two to three hours a day sifting through information using an artificial neural network to develop a program that will automatically classify the light curves of eclipsing binary stars.

An artificial neural network is a type of artificial intelligence that mimics the behaviors and functions of human brains. In other words, by using the system, Leaveck is teaching the computer to “think” like a human.

Dr. Shaukat Goderya, director of Tarleton’s astronomy program, has worked closely with Leaveck over the past year.

“These light curves are different,” Goderya said as he described the program’s creation that was illustrated on a computer screen. “We can see that. What we’re trying to do is use mathematical tools to help the computer ‘see’ the differences, much the way the human eye can.”

Leaveck explains further.

“We’re using Fourier Analysis as a way of mathematically describing the shapes of the light curves,” she said. “The way I explain this to people is that computers think in numbers, so we need to be able to describe these shapes in terms of numbers.”

The ultimate goal is to train the neural network to automatically classify the star systems on a massive scale. The project is still in its early stages, but when finished, a process that once took scientists days to complete can be accomplished in a matter of minutes.

“When it is finished, it will remove most of the routine work involved with classifying eclipsing binary stars,” Goderya said. “This is a unique and new research-there are only two or three other groups in the country working on this type of research.”

Tarleton is often at the forefront of scientific research, a testimony to the university’s commitment to excel in scholarship, teaching and learning.

“Tarleton students have won first, second and third place at national scientific research meetings such as the Sigma Xi Student Research Symposium,” said Dr. Bert Little, Associate Vice President for Academic Research. “One stu