Abstract: An estimation of sensitivity of a new measure of consistency of interval pairwise comparison matrix, i.e. interval spectral coefficient of consistency to fuzzy fundamental scales which are the most often used in fuzzy AHP methods is carried out. This coefficient of consistency may be more valid for solving some foresight problems than other known indexes and methods. The coefficient of consistency is theoretical, but not empiric attribute of consistency of pairwise comparison matrix in the sense that on determination of consistency are not used randomly filled pairwise comparison matrices. For determination of admissible level of inconsistency of interval pairwise comparison matrices application and detection thresholds are developed. To estimate the sensitivity of interval spectral coefficient of consistency a computer simulation study was performed. Spectral coefficient of consistency of FPCM is proposed as combination of ISCC in all  -levels using linear, multiplicative and min combination rules of the AHP. For determination of admissible level of inconsistency of FPCM necessary and sufficient conditions are proposed. The AHP method and spectral coefficient of consistency were used to evaluate critical technologies of energy conservation and power efficiency in Ukraine.

Keywords: interval pairwise comparison matrix, fuzzy pairwise comparison matrix, spectral coefficient of consistency, admissible level of inconsistency, interval weights, fuzzy fundamental scales, foresight problems.

ACM Classification Keywords: H.4.2. INFORMATION SYSTEM APPLICATION: type of system strategy

Link:

SPECTRAL COEFFICIENT OF CONSISTENCY OF FUZZY EXPERT INFORMATION AND ESTIMATION OF ITS SENSITIVITY TO FUZZY SCALES WHEN SOLVING FORESIGHT PROBLEMS

Nataliya Pankratova, Nadezhda Nedashkovskaya

http://www.foibg.com/ijitk/ijitk-vol06/ijitk06-4-p02.pdf