Convergence of inconsistency algorithms for the pairwise comparisons

被引:24
作者
Holsztynski, W [1 ]
Koczkodaj, WW [1 ]
机构
[1] LAURENTIAN UNIV,DEPT COMP SCI,SUDBURY,ON P3E 2C6,CANADA
关键词
design of algorithms; analysis of algorithms; iterative algorithms; algorithm convergence; performance evaluation; local and global triad inconsistency;
D O I
10.1016/0020-0190(96)00113-5
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
A formal proof of convergence of a class of algorithms for reducing inconsistency of painwise comparisons (pc) method is presented. The design of such algorithms is proposed. The convergence of the algorithms justifies making an inference that iterated modifications of the pc matrix made by human experts should also converge. This is instrumental for credibility of practical applications of the pc method.
引用
收藏
页码:197 / 202
页数:6
相关论文
共 10 条
[1]  
Crawford G. B., 1987, Mathematical Modelling, V9, P327, DOI 10.1016/0270-0255(87)90489-1
[2]   GENERALIZATION OF A NEW DEFINITION OF CONSISTENCY FOR PAIRWISE COMPARISONS [J].
DUSZAK, Z ;
KOCZKODAJ, WW .
INFORMATION PROCESSING LETTERS, 1994, 52 (05) :273-276
[3]  
GANTMACHER FR, 1959, APPLICATIONS THEORY
[4]   A Monte Carlo study of pairwise comparison [J].
Herman, MW ;
Koczkodaj, WW .
INFORMATION PROCESSING LETTERS, 1996, 57 (01) :25-29
[6]   A NEW DEFINITION OF CONSISTENCY OF PAIRWISE COMPARISONS [J].
KOCZKODAJ, WW .
MATHEMATICAL AND COMPUTER MODELLING, 1993, 18 (07) :79-84
[7]   Statistically accurate evidence of improved error rate by pairwise comparisons [J].
Koczkodaj, WW .
PERCEPTUAL AND MOTOR SKILLS, 1996, 82 (01) :43-48
[8]   SCALING METHOD FOR PRIORITIES IN HIERARCHICAL STRUCTURES [J].
SAATY, TL .
JOURNAL OF MATHEMATICAL PSYCHOLOGY, 1977, 15 (03) :234-281
[9]   COMPARISON OF EIGENVALUE, LOGARITHMIC LEAST-SQUARES AND LEAST-SQUARES METHODS IN ESTIMATING RATIOS [J].
SAATY, TL ;
VARGAS, LG .
MATHEMATICAL MODELLING, 1984, 5 (05) :309-324
[10]   A law of comparative judgment [J].
Thurstone, LL .
PSYCHOLOGICAL REVIEW, 1927, 34 (04) :273-286