It is very often that practitioners ask this question.
In reality there is not a 'best' method; probably it is better to ask about the 'most appropriate' method.
However, the most important aspect to consider is what the problem or scenario characteristics are, and from there choose the method, that best match them. It is just common sense.
It is easy to understand that a practitioner can apply a certain method
which is for instance the most used. However, if this method can't model a scenario as much close as possible, the result obtained is irrelevant and false, because it will be solving a problem that has no relation with the real scenario.
Based on the bold text, I developed a simple tool that can help the practitioner. It is not guarantee that the selected method replicates in full a certain scenario, but it will give the most close selection, and very important, gives as way to justify that selection
There are two parts labeled 300a and 300b. The first explains the procedure, while the second provides the tool in Excel for the practitioner to work, and without any calculation.
The additional advantage, is that the practitioner can make as many tests as wish, and that the characteristics for each scenario can be saved, and therefore it is easy to see in the future why a certain result was achieved
Dear Sir,
As per my conclusion that ,the different Multi-Criteria Decision methods present different ranking for the same problem with same multiple attributes and same multiple alternatives although same decision maker . hence there is not any best MCDM method for any particular problem.
Dear Udaykumar
You are right, from the point of view of ranking there is not a best method because any of then could be the best.
The best method, and that was my intention in publishing this paper, is which best represent a given scenario.
Respected Munier sir,
your last thought is absolutely correct. (i.e "The best method, and that was my intention in publishing this paper, is which best represent a given scenario"). Most of the methods are unable to handle Incomplete data, uncertainty in decision judgement ,imprecision data and unexpected attributes( e.g. In ongoing ICC world cup at England, I think now rain attribute is more preferable than other attributes even though due to nature of pitch ,advantage goes in favor of first batting team.
Even though Sir, fuzzy set theory is also failed to handle such cases.
Dear Udaykumar
Thank you for your comments.
However, I was not talking about uncertain data, there are several techniques to handle that, or about missing data or unexpected attributes, and I believe that Fuzzy logic is an excellent way to ameliorate those problems.
I am talking about the inability to take into account or modelling, the actual characteristics of a project, a fact that precedes data analysis.
I am referring to the following aspects, just to name a few, that most MCDM method DO NOT CONSIDER, because their inability to work with them, or to model them, for instance:
1 Projects with criteria interrelated.
2 Projects that have dependency among alternatives
3 Projects that use resources and have limits for each one, minimum or maximum
or both at the same time
4 Projects where it is necessary to consider joint ventures
5 Projects in several scenarios
6 Projects considering execution time, etc.
These are the types of factors that I was referring to.
The best MCDM method for a certain scenario is which can comply with this type of attributes, and perhaps for a different scenario there is different best method,.
Suppose for instance that you have a scenario with characteristics 2, 4 and 6, and that the DM has to chose between methods A, B and C
He finds that method A can handle only characteristic 4, method B, can handle the three of them, and method C can handle characteristics 4 and 6.
It is obvious that the DM must go for method B, which is the best, since it has the largest compliance of all characteristics.
After this selection, the DM can deal with the quality of data that you mention.
Respected Munier Sir,
Thank you for explaining in details with examples from your busy schedule.
You are most welcome, and I thank you for sharing your concerns with all of us, since questions as yours are necessary for us to think about situations, and not to adopt a method just because it is there, or because we have the software, or because another similar problem with solved with it.
- Badges
- Science topic