Yang Xiang, University of Guelph
Associate Director, Undergraduate Studies
Professor
Guelph, Ontario
yxiang@uoguelph.ca
Office:
(519) 824-4120 ext. 52824
Expert In
Bio/Research
Dr. Xiang’s research investigates how to build intelligent programs that can reason and support decision-making in uncertain environments. His work applies normative approaches to the construction of such systems, which are then guided by a normative theory such as Bayesian probability theory or ...
Click to Expand >>
Click to Expand >>
Bio/Research
Dr. Xiang’s research investigates how to build intelligent programs that can reason and support decision-making in uncertain environments. His work applies normative approaches to the construction of such systems, which are then guided by a normative theory such as Bayesian probability theory or decision theory, and subjected to formal analysis of performance and error bounds.
For efficient inference and decision making, Xiang's research focuses on graphical models, such as Bayesian networks, decomposable Markov networks, influence diagrams, Markov decision processes, and constraint networks. Xiang's work extends these graphical models from single intelligent agent-based to multi-agent-based systems, ex., multiply sectioned Bayesian networks, where multiple intelligent subsystems cooperate in problem-solving.
In recent years, Xiang's team has focused on refining graphical models with local models for tractable inference, and for learning such tractable models from data.
Click to Shrink <<
For efficient inference and decision making, Xiang's research focuses on graphical models, such as Bayesian networks, decomposable Markov networks, influence diagrams, Markov decision processes, and constraint networks. Xiang's work extends these graphical models from single intelligent agent-based to multi-agent-based systems, ex., multiply sectioned Bayesian networks, where multiple intelligent subsystems cooperate in problem-solving.
In recent years, Xiang's team has focused on refining graphical models with local models for tractable inference, and for learning such tractable models from data.
Click to Shrink <<