Sparse PSD approximation of the PSD cone
Sparse PSD approximation of the PSD cone
复制标题
PSD 锥体的稀疏 PSD 近似
DOI:
10.1007/s10107-020-01578-y
复制
发表时间:
2020
影响因子:
2.7
通讯作者:
Sun, Shengding
中科院分区:
文献类型:
--
作者:
Blekherman, Grigoriy;Dey, Santanu S.;Molinaro, Marco;Sun, Shengding
While semidefinite programming (SDP) problems are polynomially solvable in theory, it is often difficult to solve large SDP instances in practice. One technique to address this issue is to relax the global positive-semidefiniteness (PSD) constraint and only enforce PSD-ness on smallerprincipal submatrices—we call this thesparse SDP relaxation. Surprisingly, it has been observed empirically that in some cases this approach appears to produce bounds that are close to the optimal objective function value of the original SDP. In this paper, we formally attempt to compare the strength of the sparse SDP relaxation vis-à-vis the original SDP from a theoretical perspective. In order to simplify the question, we arrive at a data independent version of it, where we compare the sizes of SDP cone and the-PSD closure, which is the cone of matrices where PSD-ness is enforced on allprincipal submatrices. In particular, we investigate the question of how far a matrix of unit Frobenius norm in the-PSD closure can be from the SDP cone. We provide two incomparable upper bounds on this farthest distance as a function ofkandn. We also provide matching lower bounds, which show that the upper bounds are tight within a constant in different regimes ofkandn. Other than linear algebra techniques, we extensively use probabilistic methods to arrive at these bounds. One of the lower bounds is obtained by observing a connection between matrices in the-PSD closure and matrices satisfying the restricted isometry property.
登录
查看更多内容
影响因子:
2.7
作者:
Santanu S. Dey;M. Molinaro;Qianyi Wang
通讯作者:
Qianyi Wang
DOI:
--
发表时间:
1980
期刊:
影响因子:
--
作者:
R. Kaas;J. Buhrman
通讯作者:
J. Buhrman
影响因子:
2.7
作者:
E. Amaldi;Stefano Coniglio;Stefano Gualandi
通讯作者:
Stefano Gualandi
DOI:
--
发表时间:
2017
期刊:
影响因子:
--
作者:
Christopher Hojny;Hendrik Lüthen;M. Pfetsch
通讯作者:
M. Pfetsch
DOI:
--
发表时间:
2012
期刊:
OR
影响因子:
--
作者:
Matthias Walter
通讯作者:
Matthias Walter