Jeff Linderoth and Jim Luedtke, professors of industrial and systems engineering at the University of Wisconsin-Madison, have earned a best paper award from the journal Mathematical Programming Computation for their work outlining an open-source software package for solving vexing optimization problems.
The two longtime collaborators worked with colleagues at Argonne National Laboratory, including UW-Madison computer science alumnus Todd Munson, to create the software toolkit, called MINOTAUR, for mixed-integer nonlinear programming.
Linderoth says one key advantage of MINOTAUR is its ability to allow users to construct advanced algorithms for efficiently solving problems involving both nonlinear and discrete components.
“Allowing these two complexities together yields a powerful modeling paradigm for optimal decision-making, but one for which there was a dearth of effective algorithms and software for solving the models,” says Linderoth, the Harvey D. Spangler Professor.
MINOTAUR is currently being used to address issues such as optimal sensor placement, aircraft collision avoidance and financial portfolio construction.
The U.S. Department of Energy Office of Science’s Advanced Scientific Computing Research program sponsored the research.