CALD, Carnegie Mellon
Time and Place
Mauldin Auditorium (NSH 1305)
Robots in the real world can't plan in isolation. Instead, they need to worry about external forces which can change their environment while they're not looking. Unfortunately, reasoning exactly about external agents often results in intractable planning problems. Using the running example of teaching robots to play hide and seek with each other, I will describe several new general tools for simplifying and abstracting multi-robot planning problems. These tools include nonlinear dimensionality reduction and no-regret reasoning.
Dr. Gordon is a Research Scientist
in the Center for Automated Learning and Discovery at CMU and a Visiting
Professor in the Stanford Computer Science Department. He works on multi-robot systems, statistical
machine learning, and planning in probabilistic and adversarial domains. Before
joining CALD, Dr. Gordon worked at Burning Glass Technologies in San Diego,
where he analyzed databases of human-resources documents to discover patterns
that can inform hiring decisions. Dr.
Gordon received his B.A. in Computer Science from
For appointments, please contact Geoffrey Gordon