Abstract
that such approaches attempt to solve. To address such drawbacks, we develop customized solution approaches for highly symmetric copositive programs, which arise naturally in several contexts. For instance, symmetry
properties of combinatorial problems are typically inherited when they are addressed via copositive programming. As a result we are able to compute new bounds for crossing number instances in complete bipartite graphs.
| Original language | English |
|---|---|
| Pages (from-to) | 659-680 |
| Journal | Mathematical Programming |
| Volume | 151 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - Jul 2015 |
Fingerprint
Cite this
}
Exploiting symmetry in copositive programs via semidefinite hierarchies. / Dobre, C.; Vera Lizcano, J.C.
In: Mathematical Programming, Vol. 151, No. 2, 07.2015, p. 659-680.Research output: Contribution to journal › Article › Scientific › peer-review
TY - JOUR
T1 - Exploiting symmetry in copositive programs via semidefinite hierarchies
AU - Dobre, C.
AU - Vera Lizcano, J.C.
PY - 2015/7
Y1 - 2015/7
N2 - Copositive programming is a relative young field which has evolved into a highly active research area in mathematical optimization. An important line of research is to use semidefinite programming to approximate conic programming over the copositive cone. Two major drawbacks of this approach are the rapid growth in size of the resulting semidefinite programs, and the lack of information about the quality of the semidefinite programming approximations. These drawbacks are an inevitable consequence of the intractability of the generic problemsthat such approaches attempt to solve. To address such drawbacks, we develop customized solution approaches for highly symmetric copositive programs, which arise naturally in several contexts. For instance, symmetryproperties of combinatorial problems are typically inherited when they are addressed via copositive programming. As a result we are able to compute new bounds for crossing number instances in complete bipartite graphs.
AB - Copositive programming is a relative young field which has evolved into a highly active research area in mathematical optimization. An important line of research is to use semidefinite programming to approximate conic programming over the copositive cone. Two major drawbacks of this approach are the rapid growth in size of the resulting semidefinite programs, and the lack of information about the quality of the semidefinite programming approximations. These drawbacks are an inevitable consequence of the intractability of the generic problemsthat such approaches attempt to solve. To address such drawbacks, we develop customized solution approaches for highly symmetric copositive programs, which arise naturally in several contexts. For instance, symmetryproperties of combinatorial problems are typically inherited when they are addressed via copositive programming. As a result we are able to compute new bounds for crossing number instances in complete bipartite graphs.
U2 - 10.1007/s10107-015-0879-0
DO - 10.1007/s10107-015-0879-0
M3 - Article
VL - 151
SP - 659
EP - 680
JO - Mathematical Programming
JF - Mathematical Programming
SN - 0025-5610
IS - 2
ER -