Lata Kodali, DAC Ph.D. student in statistics

Lata Kodali looks at statistics as a great bridge between theory and application.

“It is  also a field that is applicable in a broad spectrum,” she said,  “and right now I see myself working in an industry position with a focus on research and design that also encourages creativity.”

Kodali has a bachelor’s degree from Carson-Newman University and a master’s degree from Wake Forest University, both in mathematics. Prior to her Ph.D. work, most of her experience was theoretical rather than applied.

On a recommendation by her undergraduate advisor, who was a Virginia Tech alum, Kodali applied to Virginia Tech’s Ph.D. program in statistics. She applied to a few other graduate schools as well but, she said, the department visit sealed the deal.

“Everyone was very friendly and encouraging, and there is a variety of research interests within the department,” she said. The atmosphere felt warm rather than competitive, and fellow students really are colleagues rather than competitors.”

Kodali is working in the Bayesian Visual Analytics (BaVA) research group with her advisor and DAC faculty Leanna House.

Her current research focuses on the uncertainty in interactive displays of data created from Weighted Multidimensional Scaling (WMDS). WMDS is a linear projection technique to display high-dimensional data into a two-dimensional projection.

“The problem with current displays is that there is no information included about how imperfect the two-dimensional projection is,” Kodali said. “My current project is using Bayesian modeling to find a way to quantify this information and display it within an interactive visualization to help guide analysts in their data explorations.“

Kodali’s interest in this area of research was peaked while assisting House with user studies in the introductory statistics course STAT 2004. The BaVA research group developed a program that incorporates interactivity of WMDS displays, essentially a non-traditional learning tool, to see what kind of inferences students could make about the data without using formal statistics.

“It was interesting to see how novice analysts handle such explorations when there are no numbers involved and they have complete freedom to look at whatever they would like,” she said.

Kodali’s other research interests include regression and ANOVA,  social science, economics, biology and environmental science.

She is on track to graduate in 2020.