Improvement of AHP method
Lavoslav Čaklović, Ružica Piskač, Vedran Šego
Abstract
Lavoslav Čaklović, Ružica Piskač, Vedran Šego
Abstract
Here we propose a new ranking method for AHP, Potential method, based on Graph theory approach. This method works fine in the case when AHP-matrix is not complete. It makes easy to define an index of inconsistency of given data in a very natural way. The main purpose of this article is to examine the behaviour of rank order under successive neglecting of data in AHP-matrix preserving the graph connectedness. As we will show, the same ranking order is obtained for complete and incomplete AHP-matrix.
OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Here we propose a new ranking method for AHP, Potential method, based on Graph theory approach. This method works fine in the case when AHP-matrix is not complete. It makes easy to define an index of inconsistency of given data in a very natural way. The main purpose of this article is to examine the behaviour of rank order under successive neglecting of data in AHP-matrix preserving the graph connectedness. As we will show, the same ranking order is obtained for complete and incomplete AHP-matrix.
Key concepts: Analytic hierarchy process, Social connectedness, Ranking (information retrieval), Rank (graph theory), Graph, Computer science, Graph theory, Matrix (chemical analysis)