| I. I. Boduro~glu, 1997. Scalable Massively Parallel Simplex Algorithms for BlockStructured Linear Programs. Ph.D. Thesis, GSAS, Columbia University, New York, NY. (http://www.ieor.columbia.edu/~ilkay/Boduroglu.thesis.ps) |
....of its elements equal to zero. Then, c , x and u will have to be permuted into c , x and u, respectively. Furthermore, we will use B as a basis for A and let B be some column permuted version of B. For a survey of existing applications of the BD LP and the GUB LP problems, see [3]. Note that an algorithm to detect block diagonal structure (if one exists) in large scale LP problems appears in [22] 1.3 Motivation for the Compact Tableau Simplex Algorithms Because data parallel architectures are array processors, we were motivated to come up with an array processing ....
....of its nonzeros in the kth block of m 3 elements, then so does the jth column of V . Theorem 2: Let the matrices A and B be defined as in the statement of Theorem 1. Then, Z = B Gamma1 A can be 4 For a survey of previous work on compact inverse algorithms and their implementations see [3]. 5 See equation (1) for the definitions of T and R. 6 See [3] for the proofs. expressed in the following form: Z = B Gamma1 A = B Gamma1 A 1 A 2 = Z 1 Z 2 = Z 1 V Z 1 QA 2 , 2) where V is the m 2 Theta m 1 matrix GammaT Gamma1 R whose columns have the same ....
[Article contains additional citation context not shown here]
I. I. Boduro~glu, 1997. Scalable Massively Parallel Simplex Algorithms for BlockStructured Linear Programs. Ph.D. Thesis, GSAS, Columbia University, New York, NY. (http://www.ieor.columbia.edu/~ilkay/Boduroglu.thesis.ps)
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC