Hi Mr @Pijush Kanti Dutta Pramanik

If you mean which method, I suggest using an expert system based on fuzzy network analysis method. Such a system, based on the parameters, relationships and rules set by the experts and the removal of ambiguity by the fuzzy view, can ensure that the decision made has the necessary efficiency.

Good luck

Dear Majid

Well, this is new. Since when the experts determine the parameters, relationships and rules?

The characteristics of the real problem have nothing to do with them?

Are you sure that the fuzzy view can remove ambiguities?

Ans what is a measure for the efficiency?

Because for that you need a yardstick and nobody has it.

Dear Master and educated

Mr @Nolberto Munier

I feel that you have a little misunderstanding in reading my comments or maybe I have not been successful in expressing my views.

As I said in response to your other comment, I will provide my explanation and answer at the earliest opportunity, no doubt either this explanation will be convincing for you or I will learn some points from you. The goal of all of us is to raise the level of knowledge and resolve such ambiguities at the same time. Of course, with respect.

Good luck

Dear Mahid

It could also be that it is my fault for not interpreting you. Why to blame yourself?

Yes, I know that you said that for the first comment of yesterday, but I was not referring to it but to your comments to Pijush

Mi friend, perhaps it is I who has to learn something from you. I am not infallible and can be mistaken as anybody else.

However, I am hard to convince. When I discuss a subject with someone I give reasons, examples and rationality, and I expect the same from the colleague with whom I am discussing. Other than that any discussion is sterile.

I am not satisfied with words, which unfortunately seems to be the only language of AHP supporters use. I need facts, axioms, rationality.

Believe it or not, in the 5 years that am in RG, I never received a rebuttal of what I said in hundreds of comments and answers... It appears that AHP defenders do not have arguments to refute me. It is not a merit of me, it is simply that AHP is indefensible.

Trust me, if I am mistaken, I will recognize it publicly

I agree 100% with you that the objective of this forum is to raise the level of knowledge, however I can't understand when you say' resolve such ambiguities at the same time'

In all honesty, this is an example of what I say about using only words. What that sentence mean?

Dear Prof. Nolberto and Mr. Majid

Thank you for your comments and valuable discussions. Sure, the complexity will be dependent on the number of alternatives and criteria and the types and interdependencies of the criteria. But I wanted to know, in general, which MCDM methods have the best run-time complexity (i.e. the program for which method takes minimum time to run on a computer), considering the number of alternatives and criteria are same for each method and there is no ambiguity and interdependencies among the criteria. In my problem, there are around 50 alternatives and at least 15 criteria (10 profit, 5 cost).

Regards,

Pijush

Dear Pijush

I believe that I misunderstood your question

For large projects I have only experience in SIMUS.

A matrix of say 120 alternatives and say 30 criteria MAY take in my laptop about 15 minutes.

I am assuming that the matrix has already electronically loaded in SIMUS, and this is the time for normalization, solving, and storing result in an Excel file.

A problem of your dimensions will take probably 4 to 5 minutes

SIMUS is in Visual Basic although it has also been written in Python and Java

If you need more information, don't hesitate in contacting me by RG or by email

[email protected], I will be glad to help

Dear Prof. Nolberto,

Thank you for your in-detail response.

Sir, actually I'm looking for some MCDM method to run for a real-time application which will be used to select some services in real time, based on the available services (alternatives) at that particular time. Here, the run time of the MCDM or the selection program should be within a few milliseconds.

Is there any MCDM method available for this type of dynamic and time-constraint environment? And if not, would minimising the number of criteria and alternatives help?

Best regards.

Dear Pijush

Well unfortunately you did not say this at the beginning, and then I assumed that you were talking of a common project.

I don't know which MCDM, if it exists, can do the job you are asking

Regarding your question, I really don't know the answer, and yes, probably decreasing the number of criteria and alternatives will help

Mr Nolberto Munier

My answer was about an expert system. Yes, in an expert system we get help from experts to know the effective parameters and the main problem options. Open questionnaires, interviews or other similar methods can be used to do this. The next steps, which are more computational, can be done in methods such as AHP, ANP and TOPSIS. And the output of this section, which are the most important parameters and options in the hypothetical problem, together with the rules obtained from the opinions of experts, form the structure of the hypothetical expert system. This is a theory that I have used in several research projects and fortunately I have received acceptable results.

And again, thank you for your constructive comments and views Good luck

Dear Pijush Kanti Dutta Pramanik

To a large extent, your research problem is similar to my dissertation. I suggest using fuzzy network analysis method to prioritize the effective parameters in the problem and decision options. Why does this method consider both external and internal communication and because of the use of fuzzy theory, the ambiguity in the opinions of people completing the pairwise comparison questionnaires is removed. Note that these questionnaires are generally answered with verbal variables. I also suggest that you use the output of this section as a structural framework for an expert system to further enrich your research.

Good luck

Dear Majid

Please accept my apology since you used the sentence expert systems from the very beginning. I misread it and assumed that you were based only on experts.

Nevertheless, I don't think that you can get reliable information about criteria importance using AHP and ANP, although most probably you can get valuable results using TOPSIS and working with entropy weights. Of course, it is only a suggestion.

Thank you for your clarification

Dear Majid Jangizehi,

Thanks for your suggestion. In my problem, the criteria and the corresponding weights are set purely based on technical specifications rather than experts' opinions and they all have crisp values. So, I'm not sure if Fuzzy would help here. My priority is to have minimal running time rather than to find a robust ranking.