|
Michael C. Ferris
Professor of Computer Sciences and Industrial and Systems Engineering |
|
Research Interests:
Algorithms, environments, theory and applications of optimization.
Research Summary
The major thrust of my research is in providing tools to
operations researchers, economists and engineers that help them to
understand and solve their models. Problem analysis routines, data
manipulations and solution visualization procedures are being
developed to aid both solver routines and model developers. I am
developing software that converts nonlinear programs into
complementarity problems without human interaction, using automatic
differentiation techniques. Current work is investigating how to
easily formulate and solve practical optimization problems using the
Condor system over a confederation of workstations.
I am interested in using optimization in applications ranging from
video-on-demand to
radiation therapy.
I also continue to investigate
robust methods for solving large-scale variational inequality and
nonlinear programming problems with applications to problems in
economics and engineering.
Algorithmic work related to complementarity solvers is attempting to
solve large scale models in a robust and numerically stable fashion.
To this end, we continue to develop new algorithms and
implementations, along with testing beds and techniques for succinctly
modeling both complementarity problems and mathematical programs with
equilibrium constraints.
Recent talks
Software
Optimization within Gamma Knife Radiosurgery
|