Say I have two matrices of points, A and B, that contain point coordinates. I want to find the point pairs that minimize the sum of euclidean differences between points pairs.
For instance, in the 1-D case, I have: A = [4 1 1.5]; B = [4 1.2 0]; If the algorithm matches the nearest pairs first (like this one), this could return pairs [4 4], [1 1.2], [1.5 0]. This would give a total difference of 1.5+.2+0 = 1.7.
I'm looking for a solution that would minimize the total difference between pairs, which gives the solution [4 4], [1 0], [1.5 1.2], for a total difference of .3+1+0 = 1.3.
This is for 10k-100k points in 3D.
Thanks for your help!