IIRC, the best thing to do was to apply clustering to the rows/columns (k-means and hierarchical both worked, but I think it was hard to find the appropriate tr

IIRC, the best thing to do was to apply clustering to the rows/columns (k-means and hierarchical both worked, but I think it was hard to find the appropriate tree height for hierarchical clustering) and then to apply seriation: http://innar.com/Liiv_Seriation.pdf http://innar.com/Liiv_Seriation.pdf