Students Present Research at International Conference
Sitting in their analytics courses last year, a group of School of Management students hardly expected their research projects to lead them to presenting at an international conference. But that’s exactly what happened for MS-marketing analytics student Jamie Wang, and MS-business analytics students Ziru Wang, Yanjing Wang, and Maryam Ahmadi this fall, as all of them took part in the 2023 INFORMS conference in Phoenix, Arizona.
Jamie, Ziru, and Yanjing, who presented together, were encouraged to look into the conference by SOM Professor Jin Fang. She recommended the students submit their course research project to the conference as a possible seminar for attendees. The three had previously heard of INFORMS while attending the spring Northeast Decision Sciences Institute (NEDSI) conference in Washington DC, where the trio won first place in the graduate students poster competition for their research focused on coronary heart disease analysis. They developed a machine learning model using patients’ medical health reports to predict the probability of the disease occurring in the next ten years.
“It was really exciting to win the competition at NEDSI,” Yanjing said. “And it motivated us to continue our research and present at INFORMS. We were thrilled to be selected to participate.”
INFORMS is one of the world’s largest conferences on data science and related fields. It brings students and industry experts from around the globe together to share knowledge and research, with a focus on allowing students to be part of the presentation experience. Jamie, Ziru, and Yanjing presented their research titled, “AI Trend in the Exploration of Public Opinion Analysis About ChatGPT.” The students used Python to analyze millions of tweets about Chat GPT, organized the data into topics using a method called Latent Dirichlet Allocation, and then figured out people’s feelings toward Chat GPT using a sentiment analysis tool. To the students’ delight, the topic brought a packed room.
“Our research project was in the ‘data mining’ track, so most of the audience were people interested in the AI industry or had expertise in social media impact research. Many provided us valuable insight and feedback, which was excellent,” Yanjing added.
Maryam Ahmadi, a second year MSBA student, presented solo at the conference. She learned about the opportunity from a faculty member as well, Professor Hamid Ahady Dolatsara.
“It was always on my mind since Professor Hamid informed us about it during our visualization class two semesters ago,” Maryam stated. “It was exciting to finally be there!”
Maryam, who also had a poster on display detailing her research on “improving urban sustainability through solar panel identification,” was lauded for her paper presentation titled, “ML-CTG: Machine Learning for Cardiotocography-Based Health Classification.” The research was influenced by concepts and skills she acquired in her machine learning class and her advanced big data computing and programming class. It brought a variety of experts and researchers in the field to her session.
“My audience was a diverse group of professionals, researchers, and students interested in the fields of operations research and analytics. There were Ph.D. students and faculty members from Beijing Institute of Technology in Beijing, China, Johns Hopkins University, University of Texas-Austin, and a number of other large universities,” she noted.
All of the students credit their coursework and experience in the School of Management for providing the skills they needed to both complete their research and communicate the results, and are grateful for the professors who have supported them on their journeys.
“Courses like Visualization Analytics and Business Intelligence and Analytics Programming had us performing analyses by preprocessing the data, filtering the unnecessary part of the mess data, and then visualizing it to make it easier to understand and present to the public,” Yanjing said.
In December, Maryam’s latest work will be presented by a colleague at the American Geophysical Union 2023 Conference. The title of her poster, “Assessing Performance of a Foundation Model for Multi-Label Image Classification of Satellite Imagery” is an expanded version of a topic she had previously worked on as a research assistant in Clark labs during the summer.
You can find these excellent students on LinkedIn: