534 Interdisciplinary Science & Engineering Complex (ISEC)
805 Columbus Avenue
Boston, MA 02120
- PhD in Computer Science, Bar-Ilan University, Israel
- MSc in Computer Science, The Open University, Israel
- BA in Computer Science, The Open University, Israel
Roi Yehoshua is a postdoctoral researcher at Northeastern University, working with Professor Christopher Amato. His research interests include deep reinforcement learning and multi-agent systems. Roi received his PhD in Computer Science from Bar-Ilan University, Israel in 2018, where he was advised by Professor Gal A. Kaminka and Dr. Noa Agmon.
- Hometown: Tel Aviv, Israel
- Field of Study: Artificial Intelligence, Machine Learning, Robotics
What are the specifics of your graduate education (thus far)?
During my PhD studies I have tackled a fundamental problem in robotics, namely the robotic coverage problem, in which a team of robots needs to visit every point in a given area. I opened a new research area by introducing an adversary model into the problem, in which the robots need to cover an area containing threats that might harm them. This problem of adversarial coverage has many real-world applications, from searching for melted fuel rods in damaged nuclear power plants to demining and surveillance of enemy forces in the battlefield.
While searching for an optimal solution to the problem, I have made an important contribution to the field by introducing a first MDP (Markov Decision Process) approach to the problem, and suggesting a new general method for solving MDP problems, that shows significant improvement over the state-of-the-art methods.
The results of my work have been published in the top relevant AI and robotics venues.
What are your research interests in a bit more detail? Is your current academic/research path what you always had in mind for yourself, or has it evolved somewhat? If so, how/why?
My research goal is to help design the next generation of intelligent robots and autonomous agents, capable of adapting and learning in complex, uncertain, and adversarial environments. I am also interested in exploring teamwork and coordination in multi-robot systems, in which robots should work together to accomplish a given task (e.g., cover a given area).
What’s one problem you’d like to solve with your research/work?
Find efficient methods for a team of robots to learn to cooperate with each other in solving complex tasks, without having prior knowledge of their environment and their other teammates’ behaviors.
What aspect of what you do is most interesting/fascinating to you? What aspects of your research (findings, angles, problems you’re solving) might surprise others?
The most interesting aspect in my work is its interdisciplinary nature, and the ability to combine theories and ideas from numerous research fields, such as machine learning, robotics, game theory, control theory, algorithms, geometry, etc., in order to find the most suitable solution for a given problem.
What are your research/career goals, going forward?
My aspiration is to build on my emerging presence as a researcher in the AI community, and make significant contributions to the field. Consequently, I hope to become a faculty member at a leading research university like Northeastern.