I am in the process of writing my master's thesis, and have stumbled (as one often do..) upon an area of statistics that I am not familiar with. My setup is, that I initially have 2 features which I cluster based on some algorithm, to find groups of companies. Hereafter, I'd like to go from an "unordered/equally weighted list" to a list wherein each company is ranked / weighted differently compared to how close the remaining companies are, to the company which I excluded (company A in picture). I have tried to visualise my thought process below:
For the secondary features, I'd use a sum of the absolute rank differences (SARD) approach, to make a "ranked" list in terms of how close the companies in Cluster 1 are to company A. However, the last step of my analysis is to assign a weight to company [B, C, D, E..] and take a weighted average. My issue is, that I have no experience in how weights can be found, nor do I know where to look in terms of literature.
Can someone help me? It would be greatly appreciated! :)
