R. G. Bland, New nite pivoting rules for the simplex method, Math. Operations Research 2 (1977), 103-107.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....which is subexponential in the dimension d and the number n of facets. Cubes have come into focus in the context of the Simplex Algorithm because the examples of linear programs that fool some of the most famous pivot rules are deformed cubes (notably the Klee Minty cubes against Bland s rule [2] and Dantzig s largest coecient rule) Moreover, the existence of a polynomial pivot rule for combinatorial cubes (i.e. polytopes whose face lattices are isomorphic to the face lattices of cubes) is as unsettled as for general linear programs. The goal of this paper is on the one hand to ....

....from a geometric realization of the d cube along with some linear function the expected number of pivot steps taken by RF (starting at an arbitrary vertex) will be at most const d 2 . A third rule which we consider in our computational study is a randomized version of Bland s rule. Bland [2] proposed the following (deterministic) rule: Before you start the Simplex Algorithm number the facets of the polytope in an arbitrary order. Within the algorithm, when you have to decide to which of the better neighbors 4 w of a vertex v you want to proceed, take the one with the property that ....

R. G. Bland, New nite pivoting rules for the simplex method, Math. Operations Research 2 (1977), 103-107.

....behaviour of QP and LCP was considered already in the early years of mathematical programming (see e.g. Parsons [21] Some combinatorial type problems as circling were recognised and solved by Cottle and Dantzig [10] A lexicographic rule was used to prevent circling. Least index resolution of LP [4,5] was adapted rst for QP by Chang and Cottle [6,7] They have proved that the Dantzig Cottle [10] Lemke [18] and Keller [15] method (the last can be considered as a generalization of Van de Panne Whinston [31] method) is nite with the minimal index rule. An interesting early application of ....

.... ij if j 2 J B ; 1 if i = j; 0 otherwise; and denote for all k 2 J B t (k) k)j ) n j=1 = 8 : kj if j 2 JB ; 1 if j = k; 0 otherwise: It is well known that t (i) t 0 (k) for all i 2 JB and k = 2 J 0 B even if the two bases B and B 0 are di erent [4,24]. This result is generally refereed to as the orthogonality property of basic tableaux. Using Lemma 2.1 and orthogonality properties the following theorem can be proved, which describes the possible sign structure of basic tableaux. Its proof can be found e.g. in [16] Here ; 0 will ....

R.G.Bland, "New nite pivoting rules for the simplex method", Mathematics of Operations Research, 2 (1977) 103-107.

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