ACI Fellowship for IU faculty brings supercomputing capacity to the classroom

Ehren Newman, associate professor of Psychological and Brain Sciences at Indiana University, has used his time as an ACI Curriculum Fellow to build out increased computing capacity for P426: a capstone undergraduate lab course studying the brain and behavior of animals. “It’s meant to put a bow on everything they have learned: statistics, scientific writing, how the brain works, scientific design, all with cutting edge tools that they can take with them almost anywhere,” said Newman.

Ehren Newman, Associate professor of Psychological and Brain Sciences at Indiana UniversityHis course applies the same technology that has led to the development of self-driving cars—a form of video analysis also called computer vision, which Newman says has led to revolution in behavioral neuroscience. “Historically, we would have to measure how many times a rat would move back and forth on a line, for example. Or you would have to watch a video and label the movements across the duration,” explained Newman. “Now with machine learning, these networks can label the video for you, leading to more advanced statistical analysis of behavioral datasets.”

After training and implementation provided by the ACI Curriculum Fellows program, Newman guides group projects as they download an application using Research Desktop (RED), a virtually accessible interface for research software. Students then label data from video and process it into training datasets, which is finally applied to a machine learning network. “From there, we go through methods of refining data and tools for evaluating performance. Your first try usually only accounts for 60-80% of the variants and so students learn to find what data the model didn’t account for and iteratively refine from there—which is just like the real world of data modeling,” said Newman.

Up until recent innovations in computational power, all of the methods available to study animal movement used actual animals, something Newman does in his own lab. “In my research, we use animals to further our knowledge. I was not convinced that introductory education to this field should require the use of animals for ethical and practical reasons.” Now, Newman can meet educational goals by using videos from his own research as a data source for students.

Newman’s collaboration with Research Engagement and Integration Project Director Robert Henschel connected the course to storage resources provided by UITS Research Technologies. “Having terabytes of data available for students without extra costs for my lab and department is a great benefit of the IU High Performance Computing (HPC) systems,” said Newman.

Robert Henschel, project director for Research Engagement and Integration at UITS Research TechnologiesStill, he says the ACI Curriculum Fellowship’s value lies in the one-on-one HPC consulting opportunity. “I know how to do brain surgery, but I would not have been able to accomplish the technical demands of this course without a real person guiding me through the initial set up,” said Newman. “Henschel’s expertise working across platforms to tailor the data storage, software needs, and computational power for this course and my research going forward was the most valuable aspect of the ACI Fellowship experience for me,” he said. “Henschel’s guidance has helped me teach this marketable skill with confidence that students are getting a professional experience.”

Are you an IU professor looking to integrate advanced cyberinfrastructure into your courses?
Learn more about the ACI Curriculum Fellows Program