This problem [question]: Sort vectors according to their distance between them is about sorting vectors based on the distance between them. What about sorting vectors based on the correlation coefficient between them. What I want to do is to sort vectors based on their "importance" in the dataset. If we have N vectors then the sorting will be V1,V2,...,Vn where V1 and Vn are less correlated then the others. V1 and V2 are most correlated whith each other...and so on. I was thinking to use pearsons coefficient as a correlation coefficient. Is this possible or do you have any idea about this issue? And is there a good algorithm to do this or we should find the correlating coefficient between each pair of vectors and then find two less correlated ones and then sort all the vectors in between the two less correlated.
Thnx for reading and answering :)