So I've got ~1000 objects where at different time points I'm able to generate a proximity matrix between them. I don't have access to the actual locations of these objects, just the proximity between each pair. Also, this proximity isn't Euclidean, it's based on a heuristic so that it is in [0,1] (1 being right on top of each other, 0 being "extremely far away"). Since these measurements are very noisy, I'd like to smooth these proximity estimates across time (again, I have one proximity matrix per time point). I've done some googling around but haven't found anything that really makes use of proximity matrices rather than the actual positions. Would anyone be so kind as to make recommendations or point me in the direction of resources regarding methods that do this? Thanks!
[link][1 comment]