Students find career potential in research with faculty

There are groundbreaking studies coming out of the Department of Information Systems — such as studying the fuel burn rate of an aircraft and predicting when a patient with a history of heart failure will experience an acute or serious event.

By Jenny Keeler

There is groundbreaking research coming out of the Department of Information Systems. Research work such as studying the fuel burn rate of an aircraft and predicting when a patient with a history of heart failure will experience an acute or serious event.

Students have the opportunity to participate in research ranging from studying the fuel burn rate of an aircraft to predicting when a patient with a history of heart failure will experience an acute or serious event — and the benefits are big.

In addition to adding “research assistant” to their resume and LinkedIn profile, students might have the opportunity to be listed as a co-researcher in an academic journal, and, in the Mayo Clinic study, develop technology to save lives.

Typically, this extracurricular research is done outside of classroom hours, beginning with an email from Professor of Information Systems Asim Roy, who invites his students in the Master of Business Analytics (MS-BA) program to help him solve an issue with research.

"This problem-solving experience is very valuable in the marketplace,” Roy says. “The students can talk about these problems and how they dealt with them during a job interview."

Taking tech invention up in the air

AVIAGE SYSTEMS, a $1.3 billion, 50-50 joint venture between General Electric (GE) and Aviation Industry Corporation of China (AVIC), is funding a research project of Roys on hardware-based machine learning at the edge of the "internet of things." To test its capability, Aviage provided data on a particular problem — prediction of fuel burn rate of an airplane based on weather conditions and other aircraft and engine parameters. AVIAGE SYSTEMS is evaluating new technologies that can be used to improve operational efficiency of airlines.

Using the Aviage data, Roy set up a competition for MS-BA students so that they can learn about solving these kinds of problems. However, he had them use other machine-learning platforms and algorithms. Three student teams participated in this competition and researched the design of an aircraft, the factors that control its speed, and the intricacies involved in operating an airplane in flight. Drs. Savio Chau and Guillermo Gallardo of AVIAGE SYSTEMS were the judges of the competition.

Sarath Chandra Raviprolu was on the winning team. Having no prior experience working with machine learning, he had to push himself out of his comfort zone to build different algorithms that could accurately predict the fuel burn rate of an airplane.

"This project manifested as one of the best opportunities to apply the learned skills from the MS-BA program,” Raviprolu says. “It was a stepping stone toward discovering my potential in this field."

Raviprolu says he had the opportunity to interact with subject matter experts, airplane pilots, and NASA engineers with the Shanghai-based company that opened a manufacturing-research hub in Peoria, Arizona, in 2015. At the end of the project, he could discuss aviation confidently.

"I absolutely believe this project made me a better person than I was before starting it," says Raviprolu, who believes the experience was invaluable and credits Roy, or the "algorithm guru," as he is known, for a meaningful learning experience. "I started with no prior hands-on work experience whatsoever with machine learning to winning the competition. It was truly an amazing journey."

Savio Chau, principal system engineer with AVIAGE SYSTEMS, says it’s a mutually beneficial relationship. “Our engineers have a chance to work with academics and students who have the knowledge of the latest technology research advances, and the open innovation environment in universities also brings impetus to the project,” he says. "Additionally, the students have the opportunity to engage with the real problems faced by the industry."

Suchismita Sanyal, who was also on the winning team, now eats, sleeps, and breathes data mining. In the research work, she performed exploratory data analysis and wrote code for machine-learning models, among other jobs. Sanyal says it was tough to fit in the research work out of the classroom in addition to her regular coursework and projects. She and her team met on weekends and spent many late nights working.

"The project was critical and interesting," says Sanyal, who graduated in May with her MS-BA. “It was a perfect opportunity to turn my knowledge into data-science experience.”

The marketplace values that type of learning experience, Roy says. "The students get a more in-depth understanding of machine learning and how to apply the ideas in different situations."

Some of the research also requires critical thinking and problem-solving. What approach to take or what model to build is not always obvious in real-life problems, Roy says. Such was the case in the Mayo Clinic research.

Detecting heart failure with data

During his time as an ASU and Mayo Clinic Fellow, Roy was presented with a problem: how to use the streaming data to predict in near real time the onset of repeat heart failure in patients, which was an unsolved issue for Mayo cardiologists.

A lack of data was not the problem. The heart patients were wearing a remote monitoring device that continually captured data about heart rate, breathing, and other biometrics. The doctors just didn’t know how to use that data to predict the onset of heart failure.

Roy got his students involved to look at all the data collected. Rather than collect and analyze a huge amount of data from many patients, and create one prediction model for all, Roy and the team decided it would be more effective to look at the specific data from one patient and make predictions based on their individualized information.

The usual process is to collect data from many patients — sometimes thousands — then build a single model and apply it to all patients.

Students are still conducting research and testing the individualized approach. “It has worked out well and has the potential to impact medicine in a big way,” Roy says.

Other benefits of undergraduate research

Recent MS-BA graduate Tapan Shah worked with Roy on a similar research project, predicting machine failure in airplane engines. He says the work gave him a better understanding of machine learning and data science and how it can be applied to business problems.

Shah also got the opportunity to hone his communication and presentation skills. He learned how to speak with technical experts and how to, in turn, explain his findings to a layperson. “I have learned many things from this research work,” he says. "The research work gave me a good understanding of machine learning and data science outside of my class."

Recent MS-BA grad Xiangjing Chen got to add the title research assistant to her resume after her work on the paper, "Sharing Product Improvement Efforts: Impact of Purchase Price and Commitment Sequence." Chen partnered with Associate Professor of Supply Chain Management Yimin Wang on the research paper, which focused on utilizing knowledge of the game theory and calculus. Chen says she analyzed the equilibrium points of two parties’ improvement efforts in different situations and provided mathematical proofs for propositions. The work strengthened her mathematical skills and gave her experience with high-level economics and supply chain management.

"Because of this experience, I finally understand that I am truly interested in doing research and would love to make it my career in the future," says Chen, who will pursue her newly realized passion and join the PhD program in the fall to continue researching. "I believe it is very beneficial to do research with professors because you can apply what you have learned in classes to the real study, and you can learn much more through practice."

Latest news