Math and Computing Research Discovery Days

While consensus exists on the importance of encryption to ensure users’ privacy, encrypted traffic poses challenges for traffic classification; an indispensable first step for many network operations carried out by network service providers. A plethora of machine learning models, including deep learners, have been suggested in the literature for encrypted traffic classification, most of which assume access to a labeled training dataset and close training and test data distributions. In this talk, I will present the challenges that an operator will face while working with a model built under those assumptions. I will then present an alternative solution based on few-shot learning. We will see how few-shot learning reduces the number of labeled data samples required for classifying a new dataset but falls short of training a generalizable learner.

Aida Sheshbolouki is a Ph.D. candidate and a member of the Data Systems Group at the University of Waterloo under the supervision of Prof. M. Tamer Ozsu. Her research involves steps towards data-driven algorithm design for explainable and interpretable analytics over streaming graphs with a particular focus on (2,2)-bicliques (known as butterflies) in bipartite structures.

Anita Layton is the Canada 150 Research Chair in Mathematical Biology and Medicine, and Professor of Applied Mathematics, Computer Science, Pharmacy and Biology at the University of Waterloo. She leads a diverse and interdisciplinary team of researchers who use computational modeling tools to better understand aspects of health and disease. Mathematics is their microscope! The Layton group collaborate with physiologists, biomedical engineers, and clinicians to formulate detailed models of cellular and organ function. Model simulations and predictions are used to answer questions such as: How should anti-hypertensive drugs be prescribed differently for men and women? Is it better to take one's medication in the morning or at night (or does it matter), given the interactions between the drug and our body's circadian rhythms?

Professor Layton is a Fellow of the Association for Women in Mathematics (2022), winner of the Krieger-Nelson Prize (2021), and winner of Canada’s Most Powerful Women: Top 100 Award (2021). She is an Associate Editor of SIAM Review Book Section, an Associate Editor of SIAM Journal on Applied Dynamical Systems, a Section Editor (AI/Machine Learning) of Hypertension, and an Associate Editor of Maple Transactions. Additionally, she serves as the Associate Dean, Research and International, for the Faculty of Mathematics, and chairs the Research Equity, Diversity and Inclusion Council at the University of Waterloo.

Yue Lyu is a Master’s student in Human-Computer Interaction at the WatVis Lab in the School of Computer Science, University of Waterloo, under the supervision of Drs. Jian Zhao and Keiko Katsuragawa. Her research focuses on building new user experiences and systems for autistic children by using advanced computer technologies to offer various knowledge and skills presentations as well as create opportunities for children to apply the skills in both the digital and physical worlds. Leveraging her background in computer science and business administration, her work aims to create solutions that are not only innovative but also practical and beneficial to the community. Yue has extensive experience in system design, database design, machine learning engineering, and mobile application development with multiple programming languages and platforms.