A STUDY OF INDICATORS FOR IDENTIFYING ZERO VARIABLES IN INTERIOR-POINT METHODS

被引:45
作者
ELBAKRY, AS
TAPIA, RA
ZHANG, Y
机构
[1] RICE UNIV CTRE RES PARALLEL COMPUTAT,HOUSTON,TX 77521
[2] UNIV MARYLAND BALTIMORE CTY,DEPT MATH & STAT,BALTIMORE,MD
[3] RICE UNIV,CTR RES PARALLEL COMPUTAT,HOUSTON,TX 77521
关键词
INTERIOR-POINT METHODS; INDICATOR FUNCTION; IDENTIFYING ZERO VARIABLES;
D O I
10.1137/1036003
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
This study is concerned with constrained optimization problems where the only inequality constraints are nonnegativity constraints on the variables. In these problems, the ability to identify zero variables (binding constraints) early on in an iterative method is of considerable value and can be used to computational advantage. This work first gives a formal presentation of the notion of indicators for identifying zero variables, and then studies various indicators proposed in the literature for use with interior-point methods for linear programming. Both theory and experimentation are presented that speak strongly against the use of the variables as indicators; perhaps the most frequently used indicator in the literature. This study implies that an indicator proposed by Tapia is particularly effective in the context of primal-dual interior-point methods. The local rate of convergence for several indicators is also studied.
引用
收藏
页码:45 / 72
页数:28
相关论文
共 51 条
[11]  
ELBAKRY AS, UNPUB NUMERICAL COMP
[12]  
GAY DM, 1991, 5TH P MEX US WORKSH, V47
[13]   ON PROJECTED NEWTON BARRIER METHODS FOR LINEAR-PROGRAMMING AND AN EQUIVALENCE TO KARMARKAR PROJECTIVE METHOD [J].
GILL, PE ;
MURRAY, W ;
SAUNDERS, MA ;
TOMLIN, JA ;
WRIGHT, MH .
MATHEMATICAL PROGRAMMING, 1986, 36 (02) :183-209
[14]   CUTTING PLANES AND COLUMN GENERATION TECHNIQUES WITH THE PROJECTIVE ALGORITHM [J].
GOFFIN, JL ;
VIAL, JP .
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS, 1990, 65 (03) :409-429
[15]   PATH-FOLLOWING METHODS FOR LINEAR-PROGRAMMING [J].
GONZAGA, CC .
SIAM REVIEW, 1992, 34 (02) :167-224
[16]  
GULER O, 1991, CONVERGENCE BEHAVIOR
[17]  
GULER O, 1990, EXISTENCE INTERIOR P
[18]   ROBUSTNESS AND NONDEGENERATENESS FOR LINEAR COMPLEMENTARITY-PROBLEMS [J].
JANSEN, MJM ;
TIJS, SH .
MATHEMATICAL PROGRAMMING, 1987, 37 (03) :293-308
[19]   COMPUTATIONAL RESULTS OF AN INTERIOR POINT ALGORITHM FOR LARGE-SCALE LINEAR-PROGRAMMING [J].
KARMARKAR, NK ;
RAMAKRISHNAN, KG .
MATHEMATICAL PROGRAMMING, 1991, 52 (03) :555-586
[20]  
Kojima M., 1989, PROGR MATH PROGRAMMI, P29