Manhattan University hosted all summer research scholars to
present their work in Kelly Commons on Sept. 26.
@MANHATTANEDU / INSTAGRAM
By Julia Ciampa, Staff Writer
On Sept. 26, Manhattan University’s Summer Research Scholars congregated in Kelly Commons to present at the annual Summer Research Scholars Program Showcase.
Over the summer, a diverse group of students from different majors, ranging from computer science to biology, were granted the opportunity to examine different unique academic interests to contribute innovative findings to the vast world of research. After creating and conducting their research, students were required to formulate a visual and oral poster presentation on the relevance of their experimental outcomes.
Advised by Kashiffudin Qazi, Ph.D., senior computer science major Tahiya Azad was eager to join the research program since her freshman year. Azad’s research titled “Automated Neural Network Selection for Workload Predictions In The Cloud” required a plethora of previous knowledge prior to conducting the research.
Azad’s groundwork tested the role that cloud providers play while leasing machines to computer users and examined the prediction of the central processing unit (CPU) and memory usage of virtual machines leased to users. The hypothesis of her project was that no single neural network model would work best for all workloads as the model with the lowest root-mean-square deviation (RMSE) was chosen for each workload and followed by a test RMSE being calculated.
This research program requires a strong academic background, as Azad emphasized. The support the program provided her will led her to improve her professional outlook on the field.
“I always wanted to do research since freshman year, but I waited to gain more knowledge about the idea of the project,” Azad said. “At the end of my junior year, I decided to begin to consider my summer research topic. This program has granted me the confidence so that I could present my work and answer the questions that the audience had, while expanding upon my academic knowledge with the help of professor Qazi. I would like to thank Dr. Qazi for all of his guidance and mentorship throughout the entire project, as he helped me greatly.”
Another computer science summer research scholar, Thomures Momenpour focused his research on “Optimizing CNN-Based Diagnosis of Knee Osteoarthritis: Enhancing Model Accuracy with CleanLab Relabeling.”
Momenpour spent his summer researching and developing a convolutional neural network (CNN) model to detect the presence and severity of knee osteoarthritis from input x-ray images. Momenpour’s research findings validated the efficacy of deep learning in medical imaging by presenting a promising framework for advancing knee osteoarthritis (KOA) diagnosis while showcasing the potential of CleanLab for their labeling of datasets.
Momenpour went over the meticulous challenges he faced while conducting his experiment.
“My research work was undoubtedly rewarding, yet rigorous,” Momenpour said. “Personally, I would say the biggest challenge was getting the data I collected to be clean enough to be used for the model. This was extremely tedious, as the data set originally had a lot of background noise in it, or room for error, which means the x-ray images I took might not have been correctly labeled, which was the greatest challenge I faced.”
One thing both scholars had in common was their research advisor, Adriane Bilous, Ph.D.
“The most rewarding part of assisting with this program is watching the students fall in love with research,” Bilous said. “Everyone comes in with an idea of what they’re going to be when they grow up. Seeing them find something they love to do and all of a sudden they fall in love with research is extremely rewarding. These students are so talented, and finding what they love to do right now is steering them towards whatever that next step is, and seeing them exceed is so special.”
As the scholars celebrated the finalization of their research, Momenpour reflected on his favorite part of the program.
“My favorite part was the independent work and how we were trusted enough to learn the material by ourselves,” Momenpour said. “The research we conduct is a real world problem that we can contribute our knowledge to, rather than theoretical problems that we face in the classroom.”
Azad encourages prospective students to take advantage of this opportunity, and allow themselves to learn a new, interactive form of education.
“As a proud research scholar, I would definitely recommend future students to do this program, and allow themselves to learn from their advisors,” Azad said. “Your academic advisors are always there to guide you through research so I’d say if you have the chance to do this program, just go for it!”