Hi all
In AHP my approach is to perform all the calculation (inc Consistency check) of each decision maker (DM). Only then the weight result from all DM averaged using geomean. I reckon that is more sensible rather that average the pairwise score of all DM prior to do the AHP calculation.
But then how about in Fuzzy AHP (FAHP)?. Because i saw some paper average the TFN result from all DM first prior to do all the weight calculation.
Can anybody give insights? Thanks
Dear Elia Oey,
I suggest you to see links and attched file in topic.
- FUZZY ANALYTIC HIERARCHICAL PROCESS TO DETERMINE THE ...
https://www.ijma.info/index.php/ijma/article/view/2238/1361
- Multiple Attribute Decision Making: Methods and Applications
https://books.google.dz/books?isbn=1439861587
- a note on multi-criteria inventory classification using ... - doiSerbia
www.doiserbia.nb.rs/img/doi/0354.../0354-02431002293R.pdf
Best regards
Dear Elia Oey,
It's possible to use geometric average for calculation average TFN a prior to do all weights calculation. Please see following papers:
Article A fuzzy supplier selection model with the consideration of b...
Article An integrated fuzzy AHP and TOPSIS model for supplier evaluation
Best regards
Dear Elia
I agree with you about what method you prefer.
If I may I would like to suggest considering only their pair-wise comparisons as they come, that is paying no attention to inconsistencies, which are only useful if you use Eigen Value method, and of course using the geo to find the average
The reason for not considering the inconsistencies is that you can work with more realistic values, even based on preferences, because they are not adjusted to the transitivity condition of less than 10 %
In addition, if you have more that 3 alternatives, the Eigen Method is not as accurate as the geo
- Badges
- Science method
- Similar topics
- Methodology