Biomimetic model for computing missing data imputation and inconsistency reduction in pairwise comparisons matrices.

Abstract

A biomimetic model is presented to compute missing data imputation and reduce inconsistencies in pairwise comparisons matrices. The proposed regeneration method emulates three primary phases of a biological process: identifying the most damaged areas (by identifying inconsistencies in the pairwise comparison matrix), cell proliferation (filling in missing data), and stabilization (optimization of global consistency). An iterative algorithm is employed to correct inconsistencies and compute missing data imputations within the pairwise comparison matrix. The results demonstrate that the biomimetic approach is robust and reliably converges to a consistent solution.