Belabbas researches control theory and its applications

ECE Assistant Professor Mohamed Ali Belabbas joined ECE in the fall of 2012. He comes to Illinois from a postdoctoral position at Harvard where he had been since receiving his PhD in 2006.

Belabbas, a researcher in the Coordinated Science Lab, is an expert in the area of control theory, particularly decentralized control. His research examines what different agents can achieve as they experience differing amounts of cooperation and control.

“Cooperation is a fundamental tool, used by unicellular organisms, by the machines we build, and by us every single day. The fundamental questions you want to answer there are: given certain amounts of cooperation, what can the agents achieve?” explained Belabbas. “If everybody [in a group] has a completely selfish behavior, you can imagine the group will not be able to achieve much. If they cooperate more, the group will be able to achieve more. Can we quantify this rough rule of thumb? Very simple questions such as ‘What kind of cooperation is needed for a particular task’ are not yet that well understood.”

Belabbas also plans to investigate new research areas at the U of I. “I have always had a keen interests in cognitive neuroscience and psychology,” he said, “and I feel that we are getting to a point where experiments and data coming from these fields are amenable to the type of quantitative analysis engineers do best.”

For Belabbas, the long-term objective here is to understand the brain. “From a control theorist point of view, the brain is massive system where cooperation between simple units, as between neural cells, is overwhelmingly the most important part of the system,” he said. “This is the reverse of what we usually design as engineers, where often cooperation is an added value to highly sophisticated units.”

Belabbas said that he is happy to be part of the University. “The University of Illinois offers a tremendous environment for discoveries lying at the interface of two fields [engineering and cognitive sciences], not only within the College of Engineering, but also with the University as a whole and I am eager to tap into this resource,” he said.

This spring Belabbas is teaching ECE 555: Control of Stochastic Systems. In the future he would like to develop a new course in geometric control.

“I like to see teaching as an almost required part of the research process,” he said. “First, by teaching topics not directly aligned with what your current research problem is, you are forced to keep things in perspective. Second, having 15 to 60 minds in training chewing on what you tell them, you are bound to hear unusual questions and points of view, this is my favorite part of the process.”

In his previous teaching positions at Yale and Harvard, he has taught in the courses in signals and systems, decision theory and machine learning, and random processes.

Whenever possible, he likes to assign final projects instead of final exams. “Final projects are much more work for the teaching staff, but they can be much more rewarding. A group of students in my decision theory class used the tools they had just learned to enter a competition launched by an insurance company to predict hospitalizations based on past medical records. With a small team and little resources, they ranked in the top 3%, ahead of many more experienced team with more resources. This can be much more exciting than sitting 3 hours in class answering questions. Of course, not every class is as well-suited for that.”

When not conducting research or teaching, Belabbas likes to run competitively. He mostly participates in 5K and 10K races and has done a few 200 miles relay races.  He also has a passion for music, and played guitar in a group during his time in Boston. “There are quite a few venues in Champaign that have live music,” Belabbas said. “I am hoping to play more while I am here.”