Advanced Optimization Technology

Large-scale applications of optimization demand algorithms that make effective use of high-performance computing. We are using a multi-pronged approach:

Exploring new approaches to optimization algorithms, such as filter and trust-region methods.

Investigating new optimization paradigms, such as optimization problems with complementarity constraints.

Making optimization technology available to Internet users through the NEOS Guide and Server.

Designing optimization software that is interoperable with linear algebra, partial differential equations, and automatic differentiation software.

Research Projects

Toolkit for Advanced Optimization

NEOS Guide

NEOS Server

Optimization Technology Center

Investigators

Related Research Projects at Argonne

PETSc: Software for nonlinear PDEs and linear algebra

ADIFOR and ADIC : Software for computing derivatives in Fortran and C.

Educational and Employment Opportunities

2004 Summer Research Positions: Givens Associates

Undergraduate Program in Computational Science

For further information, please contact Jorge Moré.

Mathematics and Computer Science Division, Argonne National Laboratory.