קולקוויום מחלקתי 9.1.24
DR. AYAL TAITLER
University of Toronto
Will lecture on
Data Driven Decision Making: Leveraging Models for Real-World Solutions
Real-world challenges often manifest as hybrid systems, encompassing both discrete and continuous attributes, exhibiting stochastic behaviors, constraints, and the need for cooperation, or at the very least, synchronization between agents—whether they are human or artificial. Specific areas of focus encompass intelligent transportation, robotic navigation, and networked control problems, all of which demonstrate all the aforementioned characteristics, compounded by the presence of nonlinear governing dynamic models. Importantly, prior knowledge of these challenges is available in the form of models that can and should be exploited. In this talk, we will introduce these challenges and the associated complexities. We'll begin by providing a concise overview of a domain-independent descriptive language for sequential decision making problems, RDDL, and some recent extensions to it. Following that, we will unveil novel modeling and simulation tools, facilitating automatic transition from formal descriptions to functional, interactive OpenAI Gym environments, accommodating both model-free and model-based approaches. Finally, we will present a domain-independent, data-driven yet model-driven solution methodology, with the ability to tackle a wide range of problems.
Dr. Ayal Taitler is a Lyon Sachs postdoctoral fellow in the Department of Mechanical and Industrial Engineering at the University of Toronto. Ayal's primary research focus lies in the realm of hybrid discrete-continuous problems, especially when considering (abstract/partial) models. Ayal completed his Ph.D. in the Technion Autonomous Systems and Robotics interdisciplinary program. His doctoral research focused on mixed discrete-continuous planning for autonomous robotic missions. During his Ph.D., Ayal served as the lecturer, teaching the advanced control theory course in the Electrical and Computer Engineering faculty. Prior to his Ph.D., Ayal earned his Master's in reinforcement learning and Bachelor's degrees, both from the Faculty of Electrical and Computer Engineering at the Technion. Furthermore, Ayal has more than a decade of industrial experience within the Israeli defense sector and high-tech enterprises, working in software development and research roles.
ONLY VIA ZOOM