Consistency thresholds for Hierarchical Decision Model

Abstract

The objective of this research is to establish consistency thresholds linked to alpha () levels for HDM's (Hierarchical Decision Model) judgment quantification method. Measuring consistency in order to control it is a crucial and inseparable part of any AHP/HDM experiment. The researchers on the subject recommend establishing thresholds that are statistically based on hypothesis testing, and are linked to the number of decision variables and  level. Such thresholds provide the means with which to evaluate the soundness and validity of an AHP/HDM decision. The linkage of thresholds to  levels allows the decision makers to set an appropriate inconsistency tolerance compatible with the situation at hand. The measurements of judgments are unreliable in the absence of an inconsistency measure that includes acceptable limits. All of this is essential to the credibility of the entire decision making process and hence is extremely useful for practitioners and researchers alike. This research includes distribution fitting for the inconsistencies. The superb fits obtained give confidence that all the statistical inferences based on the fitted distributions accurately reflect the HDM's inconsistency measure.

Place Branding and Urban Regeneration as Dialectical Processes in Local Development Planning: A Case Study on the Western Visayas, Philippines

“…Responses were computed using the row geometric mean method (RGMM), as proposed by Golden and Wang [37]. All decisions were computed to ascertain their weight and verified based on consistency, with CR ≤ 0.2 as the tolerance value [38,39]. After performing these computations, 26 of the 30 responses (87%) were used for analysis and later compared to those from the other three groups.…”

A Proposed Feature Selection Particle Swarm Optimization Adaptation for Intelligent Logistics—A Supply Chain Backlog Elimination Framework

“…For example, the age feature could be significant to identify the chemotherapy dose, however, it could be weakly significant for the retailing field, while it could be insignificant for the real state. Identifying sixty percent acceptance for weak significance followed the research [32] which limited the threshold for ten decision variables to be at least sixty-one percent with the least variance equal to 0.1. Additionally, the least variance level is selected as the classification accuracy becomes higher with the increase of the invariance level of the variables [33].…”

“…It means that if there are ten contributing classification algorithms in the examination process, then the features subset should either have a performance value higher than the threshold for eight of these algorithms or have an average performance value higher than the threshold. This percentage is identified following the research [32] which reached a maximum of seventy-two percent of the inconsistency threshold for a variance level equal to twenty-five percent. As the classification accuracy has a negative relationship with the variance level, therefore, this percentage is identified as an exhaustive examination of the feature in the classification task to ensure its significance in the worst-case scenario.…”