Kendall’s Tau

First page Back Continue Last page Overview Text

Notes:

We have clustered among 21 similarity coefficients using Kendall’s Tau. It is a rank correlation coefficient which uses the relative ordering of ranks: higher in rank, lower in rank, or the same in rank. Here we have the formula for Kendall’s Tau.The value of Kendall’s Tau being 1 means two rankings are in perfect agreement and the value of Kendall’s Tau being –1 means two rankings are perfect disagreement. We Used Kendall’s Tau values to produce a distance matrix then cluster the coefficients. The clustering results in 13 groups, very similar to the earlier work.