On the expected condition number of linear programming problems (Q1402168): Difference between revisions
From MaRDI portal
Created claim: Wikidata QID (P12): Q57733248, #quickstatements; #temporary_batch_1711234560214 |
Normalize DOI. |
||
| (One intermediate revision by one other user not shown) | |||
| Property / DOI | |||
| Property / DOI: 10.1007/s00211-002-0385-1 / rank | |||
| Property / DBLP publication ID | |||
| Property / DBLP publication ID: journals/nm/CuckerW03 / rank | |||
Normal rank | |||
| Property / DOI | |||
| Property / DOI: 10.1007/S00211-002-0385-1 / rank | |||
Normal rank | |||
Latest revision as of 19:27, 10 December 2024
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | On the expected condition number of linear programming problems |
scientific article |
Statements
On the expected condition number of linear programming problems (English)
0 references
19 August 2003
0 references
The authors examine the condition number \(C(A)\) introduced by \textit{D. Cheung} and \textit{F. Cucker} [Math. Program. 91A, No. 1, 163-174 (2001; Zbl 1072.90564)] for linear programming. It is assumed that the matrix coefficients are independent, identically distributed standard Gaussian random variables. The moments of \(C(A)\) are estimated. When \(n\) is sufficiently larger than \(m\), when \(A\) is \(m\times n\), then \(E(\ln(C(A))= \max\{\ln m,\ln\ln n\}+ O(1)\). Bounds of an alternative condition number are also estimated.
0 references
condition number
0 references
linear programming
0 references
