- אירוע כבר עבר.
Tools and Applications for Multi-Agent Coordination
דצמבר 3, 2019 @ 12:00 pm - 1:00 pm
Speaker: Dr. Harel Yedidsion, Department of Computer Science, The University of Texas at Austin
Time: 12:00 – 13:00
Place: Building #3A, lower conference room, floor 2 (floor 1 in the elevator), Ariel University, Ariel
As multi-agent systems become more abundant and increase in scale, there is a growing need for efficient mechanisms, which determine how the agents share knowledge, coordinate and collaborate. Practical applications of multi-agent systems include autonomous vehicles, drones, service robots, sensor networks, wearable sensors, IoT devices, and last but not least, humans. Ideally, these multi-agent systems would quickly reach optimal solutions, produce strategy-proof and fair outcomes, relying on distributed decision making, using minimum communication and computation overhead. In reality, however, calculating the optimal solution is often not tractable for large scale systems that need to quickly react to changes, and distributed decision making often leads to locally optimal solutions.
In this talk, I will present a number of multi-agent applications, discuss the challenges they present, and the diverse set of tools used to design efficient solution methods, including Distributed Constraint Optimization, Computational Geometry, Game Theory, Planning, Human-Robot Interaction, and Reinforcement Learning.
Dr. Harel Yedidsion is a postdoc researcher in the Computer Science Department at the University of Texas at Austin, hosted by prof. Peter Stone, and is part of the Learning Agents Research Group. He received his Ph.D. from the Department of Industrial Engineering and Management at Ben-Gurion University and his dissertation focused on developing a framework and distributed algorithms to represent and solve distributed multi-agent coordination problems. Currently, Harel works on developing autonomous service robots, and smart transportation applications. These projects encompass many aspects of AI development including perception, navigation, planning, human-robot interaction, natural language processing, reinforcement learning, mechanism design, and fair division. Harel’s main research focus is on how to design intelligent multi-agent systems that can efficiently cooperate to solve tasks in dynamic environments.