Title
Theoretical convergence of large-step primal-dual interior point algorithms for linear programming
Abstract
Abstract: This paper proposes two sets of rules, Rule G and Rule P, for controllingstep lengths in a generic primal-dual interior point method for solving the linear programmingproblem in standard form and its dual. Theoretically, Rule G ensures the globalconvergence, while Rule P, which is a special case of Rule G, ensures the O(nL) iterationpolynomial-time computational complexity. Both rules depend only on the lengths of thesteps from the current iterates in the primal and dual spaces to the...
Year
DOI
Venue
1993
10.1007/BF01581234
Math. Program.
Keywords
Field
DocType
theoretical convergence,large step,linear program,linear programming,global convergence,poly- nomial-time convergence.,primal-dual interior point algorithm,large-step primal-dual interior point,dual space,computational complexity
Convergence (routing),Mathematical optimization,Algorithm,Line search,Duality (optimization),Linear programming,Time complexity,Iterated function,Interior point method,Mathematics,Computational complexity theory
Journal
Volume
Issue
ISSN
59
1
1436-4646
Citations 
PageRank 
References 
18
1.74
18
Authors
3
Name
Order
Citations
PageRank
Masakazu Kojima11603222.51
Nimrod Megiddo24244668.46
Shinji Mizuno3792153.37