I had this interesting conversation on using AHP, ANP and Graph theory to determine offshoring mix with my good friend Navneet over phone the other day. We had a healthy debate on futility of using graph theory to solve the problem, and some pointers to look at Influence Diagrams. I did a search up on the Web for Influence diagrams (ID) and hit upon a very interesting discussions on this and Morphological analysis (MA).
While debating about the merit Graph theory, I did realize that the challenges of applying it to solve the OOM problem. Think about it how am I going to get the strength of the edges connecting 2 nodes and the influence of the nodes itself on the OOM. Apparently, I do agree with Navneet that the OOM factors, its (inter/intra)influences and the objective function of OOM are not physical phenomena to be objectively used to reach the optimal OOM goal. But going back to using AHP for the same still does not convince me!
So the endeavor continues!!!!
I somewhat liked the idea of using the MA method. MA method is specifically meant to handle the challenge that I'm perplexed with i.e. the the complex nature of systems (OOM) and the unquantifiable nature of the factors (both nodes and edges) to determine the optimal OOM. The MA method has been used successfully in several fields: Zoology, Geology, Sociology and several other-logies. Now the challenge of MA itself is to define the problem very well, identify the influencing factors, establish the boundary conditions for each factors, Analysis-Synthesis loop to identify a consistent configuration (principles of contradiction and reduction) applying filters: logical contradictions, emprical constraints, and normative constraints.
Pretty interesting to think about the above. Now I'm more confused looking at above that the marrying of so many concepts: Exploratory analysis, Constraint theory, influence diagrams, etc. that can help in determining OOM. I'm not sure I can apply any of the above unless I get a broad set of principles based on which I can zero into a good principle.
Bloggers please help me, give me the direction.....
Subscribe to:
Post Comments (Atom)
"Now the challenge of MA itself is to define the problem very well," ... This is the key Sunil! before applying any technique formulating the problem - staying sleeping with the problem is the key... I think once done this way .... the framework starts emerging and may be techniques should emerge. Having said that, it is not so top down anyway, sometimes it makes an interesting approach to use the techniques and solve the problem - or frame the problem in techniques/framework terms... Nevertheless the journey is the reward ...anyway! :)
ReplyDelete