NSF: Interior-Point Methods for Convex and Semidefinite Programming
Project Description
Interior-point methods have proven to be efficient for the solution of
large, sparse linear and quadratic programming problems.
We propose to extend these methods to cover more general classes of problems.
In particular, we plan to apply interior-point technology to create efficient
algorithms for the
solution of convex programming problems and semidefinite programming problems.
The first class is important for general nonlinear optimization and the second
has important applications to integer programming.
This web page provides information regarding the papers, books, software, etc.
produced with the support of the NSF through grant CCR-94-083789.
3-D Models of the Optimal Solution to Some Engineering Optimization
Problems
Journal Publications
-
Extension of Piyavskii's Algorithm to Continuous Global Optimization.
R.J. Vanderbei.
J. Global Optimization. To appear 1999.
-
Symmetrization of Binary Random Variables.
C. Mallows, A. Kagan, L. Shepp, R.J. Vanderbei, Y. Vardi.
Bernoulli. To appear 1999.
-
Primal-Dual Affine-Scaling Algorithms Fail for Semidefinite
Programming.
M. Muramatsu, R.J. Vanderbei.
Mathematics of Operations Research, 24, 149-175,
(1999).
-
LOQO User's Manual--Version 3.10.
R.J. Vanderbei.
Optimization Methods and Software. To appear 1999.
-
LOQO: An Interior-Point Code for Quadratic Programming.
R.J. Vanderbei.
Optimization Methods and Software. To appear 1999.
-
An Interior-Point Method for Nonconvex Nonlinear Programming.
D. Shanno, R.J. Vanderbei.
Computational Optimization and Applications. 12, 1999.
-
The Gauss-Newton Direction in Semidefinite Programming.
Kruk, M. Muramatsu, F. Rendl, R.J. Vanderbei, H. Wolkowicz.
Submitted to SIAM J. Optimization.
Book
Software and Data
