Cross Distance - How is it calculated?

sgenzer

Question

I've ran Cross Distance using the Euclidean distance metric on two Excel files: one containing 1 row of relevant keywords etc and one containing 504 rows. The aim was to compare the similarity of each row to the relevant keywords. It all worked, which is great...except I don't know what the numbers mean! Help?

Answer

Let's say you have two attributes: age and salary. What Cross Distance is doing for each item in the reference and the request set is:

distance(ref_i, req_j)

if you use Euclidian distance you get:

d = sqrt ( (age_i - age_j)² + (salary_i - salary_j)² )

Martin

--------------------------------------------------------------------------
dr Martin Schmitz

Head of Data Science Services at RapidMiner