Led by Holger Hoos and Kevin Leyton Brown, this research group studies the empirical behaviour of algorithms, with a focus on state-of-the-art methods for solving NP-hard problems from AI and beyond. Using and improving cutting-edge methods from machine learning, optimization and statistics, research in empirical algorithmics aims to characterize and automatically improve the performance of algorithms, including those inaccessible to theoretical analysis. Work by the empirical algorithmics group at UBC/CS has lead to substantial improvements in the state of the art in solving a wide range of prominent problems, including SAT, AI Planning and Mixed Integer Programming, and won numerous awards. It has also produced methods and approaches at the forefront of research in various areas of AI and Operations Research, and it has recently given rise to Programming by Optimization (PbO), a new paradigm for the development of high-performance software.
To learn more, please see: