Decision Diagrams for Optimization

Willem-Jan van Hoeve (Carnegie Mellon University)

12 July 2023, 16:00-17:30 | Echo-Hall B1 | https://collegerama.tudelft.nl/Mediasite/Channel/eemcs-cs-distinguished-speaker-lectures-cs-dsl/watch/0d960cc1a164454f89b121e4511966381d


Abstrat

This presentation will give an introduction to the use of decision diagrams for combinatorial optimization.  It will focus on developing a novel generic branch-and-bound framework based on decision diagrams.  Working from a state-based formulation (e.g., a dynamic program), relaxed decision diagrams provide dual bounds, restricted decision diagrams provide primal bounds, and an exact search is defined based on the nodes of the diagrams.  We demonstrate that this approach outperforms or is competitive with the state of the art on various problem domains.  We will also discuss how an extension of this method called 'column elimination' can be integrated with a linear programming solver to obtain strong dual bounds for integer programming problems.

Willem-Jan van Hoeve

Willem-Jan van Hoeve is the Carnegie Bosch Professor of Operations Research and Senior Associate Dean of Education at the Tepper School of Business, Carnegie Mellon University. His research focuses on developing new methodology for mathematical optimization, data mining, and machine learning, with applications to network design, scheduling, vehicle routing, health care operations, and analytical marketing. He published over 50 academic articles, and co-authored the book "Decision Diagrams for Optimization." Van Hoeve’s research has been funded by the National Science Foundation, the Office of Naval Research, and two Google Faculty Research Awards. He has consulted for a variety of companies including FedEx Ground, Exxon Mobil, PNC Bank, Bosch/Siemens, Charter Steel, Kalibrate, and others. Van Hoeve received the Tepper School's MBA Teaching Award twice, as well as the MSBA Teaching Award.