Three Purdue professors win awards for early career achievement, innovation

WEST LAFAYETTE, Ind. — As Purdue celebrates 150 years of Giant Leaps in world-changing discovery and development, three more researchers are serving as shining examples of the university’s place as a leading intellectual center.

David Purpura, associate professor of human development and family studies; Chunyi Peng, assistant professor of computer science; and Elias Bareinboim, assistant professor of computer science, have received Early Career Development (CAREER) awards to further their research from the National Science Foundation.

The awards support the development of individual research programs of distinguished scientists early in their careers. Recipients may be involved in scientific computing, biological or environmental research, basic energy sciences, fusion energy sciences, high energy or nuclear physics. Over the next five years, the award will provide Purpura with $1,444,280, Peng with $554,995 and Bareinboim with $499,712.

Purpura researches how young children in preschool and primary school learn math and how language concepts affect that development.

Purpura’s project, “Mechanisms Underlying the Relation between Mathematical Language and Mathematical Knowledge,” will examine the process by which math language instruction improves learning of mathematics skills in order to design and translate the most effective interventions into practical classroom instruction. The first objective is to examine if quantitative and spatial math language affect the development of different aspects of mathematics performance, and the second objective is to examine how quantitative math language versus, or in combination with, numeracy instruction affects numeracy development.

Successful development of numeracy and geometry skills during preschool provides a strong foundation for later academic and career success.

“This project will allow us to better understand how early mathematics development is influenced by specific language components so that we can develop refined and targeted instructional methods and tools,” Purpura says. “Through this project, and in collaboration with a professional illustrator and professional children’s book author, we will be developing and refining several picture books for children with embedded math content and then evaluating the effects of using these books on children’s knowledge of early mathematics concepts. These books will be able to be used by parents and teachers to support their children’s learning.”

Peng’s project, “Amplifying Intelligence in Mobile Networked Systems,” aims to enhance intelligence to the 4G and 5G subsystems. Current cellular network systems remain closed to the research community and don’t expose sufficient information on their network behaviors. She will explore a novel data-driven, verifiable approach to open them up and equip with learning and reasoning capabilities. This will bring transformative innovations to current cellular network research and beyond, enabling new designs to optimize performance and enhance reliability.

Peng will also continue to work closely with mobile network companies for possible technology transfer. She will recruit and train a new generation of engineers and students, including those from under-represented groups.

“This project will open a new direction for researchers, especially academia ,to conduct mobile network research and make intellectual contributions to mobile internet infrastructure for our society,” Peng says. “The project is disruptive and also complements the ongoing efforts on wireless access and architectural innovations.”

Bareinboim researches causal interference and its applications to data-driven sciences, including health and social sciences.

His project, “Approximate Causal Interference,” seeks to develop a general algorithmic theory of approximate causal interference that is capable of producing more robust, reproducible and generalizable causal explanations. Currently, state-of-the-art machine learning methods can only be applied in large-scale settings when considering correlation, not causation, which is the ultimate goal of scientists. Knowledge available to scientists does not always match what the theory expects, and the theory does not always accept more relaxed causal specifications as input. Some researchers continue to make their claims even when the required conditions are not met.

His project will bridge the gap between the conditions entailed by the theory and the knowledge available at the hands of the scientist. He seeks to characterize the trade-off between the combination of data and background knowledge available, and the strength of newly hypothesized causal explanations. Bareinboim will construct approximation schemes allowing inputs that are coarse and imprecise while generating outputs that are still causally meaningful.

“This research will offer foundational grounding for most of the data science inferences made today, which is largely ad hoc,” Bareinboim says. “It will impact the practice of several data-intensive fields that are built on cause-and-effect relationships, including econometrics, education, bioinformatics and medicine.” 

Writer: Kelsey Schnieders Lefever, kschnied@purdue.edu 

Sources: David Purpura, purpura@purdue.edu

Elias Bareinboim, eb@purdue.edu

Chunyi Peng, chunyi@purdue.edu