ProfessorAli Bereyhiis an assistant professor, teaching stream, with the Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE). He’s been a member of the ECE community since 2023 when he joined the department’s Wireless Computing Lab as a researcher focused on distributed machine learning.
Prior to his work with ECE, Bereyhi was apostdoctoral researcher in the IdePHICS Lab atÉcole Polytechnique Fédérale de Lausanne, Switzerland, as well asat the Institute for Digital Communications (IDC) atFriedrich Alexander Universität (FAU). There he was selected as best lecturer in the Faculty of Engineering for his course lectures on information theory and coding.
His areas of expertise lie within the intersections of machine learning, statistical signal processing, information theory and statistical physics.
Could you explain the focus of your research?
I did my undergraduate degree in electrical engineering and then became interested in information theory during my graduate studies. My doctoral research focused on statistical learning and its connections to statistical physics.
Generally speaking, my research focuses on limits and the practical aspects of machine learning. Many ideas in machine learning require a high consumption of resources, and due to this, are not being deployed at large scale. In my research, I investigate designs that enable efficient implementation of these ideas with limited, and sometimes scarce, <a href="https://myappsplus.com/samsungs-new-open-galaxy-buds-on-now-available-for-purchase/” title=”Samsung's new open Galaxy Buds On now available for purchase”>available resources.
I also study the limits on the capability of machine learning algorithms to understand the power of practical artificial intelligence (AI) solutions. These are important topics in AI, as they enable us to understand the extent to which machine learning systems can be deployed, considering our available resources, and what we should expect from them in the future.
Why did you choose to work at U of T?
As someone with a relatively diverse work experience, I find the environment at U of T exceptional. The atmosphere is diverse in many respects, providing scholars with great opportunities for innovation, collaboration and professional development. U of T also offers the opportunity to work with exceptionally gifted and talented students who create meaningful impact on scientific and technological developments.
As a new professor, what’s one piece of advice you would give to new students?
I believe the time one spends as a student at university is a golden moment in their life. This is the best time to explore and learn. Much of the training we get during our time at university can come in very handy later in life when we have less time to learn new things. I think it’s best to be curious, not take anything for granted and to always ask questions. These qualities create opportunities to learn more and eventually find the way of thinking and working that fits one best.
What do you hope to accomplish during your time at U of T Engineering?
First and foremost, I hope to be the one who trains the best engineers! I consider student training my number one mission and hope that during my career I can motivate and train professionals who can serve our society. I also hope that the research accomplishments on my team can create meaningful impact and developments in the understanding of AI, its power and its limits.
What is the most memorable experience in your career so far?
During the last year of my doctoral study, I attended a talk by Professor Yan Fyodorov from the Department of Mathematics at King’s College London. Although he was working in a different field, he presented a result that was closely related to my research. After my graduation, I reached out to him to share my thoughts about the connections between what he had developed on mathematical physics and what I was exploring in my research on learning algorithms. Professor Fyodorov came back to me with very encouraging feedback that motivated me to work on the topic and develop my first research proposal and an individual PI. The outcome of the research was so well received, it received the Walter Benjamin Fellowship Award from the German Research Foundation.
The most memorable moment of the project was when I received an invitation to give a plenary talk at the Random Matrices and Random Landscapes conference in 2022 to share my findings on this topic. It was an event in honour of Professor Fyodorov’s 60th birthday. The presenters were hand-picked by the organizers, and they were all established mathematicians. Professor Fyodorov was too kind and personally asked the organizers to include me as a speaker. It was a rather unique feeling to present among several well-known mathematicians, many of them having theorems and laws named after them!
Finally, is there anything unexpected about yourself that you’d like to share?
Probably that my personal passion is history. It has been one of my interests since I was a teenager. I used to attend lectures on history while I was an undergraduate student. I don’t have any specific training in history, but I spend a lot of my free time learning about historical events and their connections to one another.
I also include this interest implicitly in my work: in my doctoral dissertation, I had a separate notes section on each chapter, where I gave a short historical note on the developments of my topic. I also incorporate historical perspectives throughout my courses! For example, in my machine learning course, I tell students how scholars have sought to understand the process of human learning, from McCulloch and Pitts trying to model our neural system computationally, to how Rosenblatt thought of perceptron as an artificial thinking machine. I feel that this makes the course more attractive too, but I guess I am assuming that everybody likes history as much as I do!
