appending together data sets corresponding to all the entities assigned to a cluster and performing a least square regression, wherein if there are m entities with n data elements each, then the model fitted in the log space is lnY=lnA+??M+??+ln??, where Y is a vector of size mn, A is a matrix of [a1 a2 . . . am], a1 being a base sale for data set 1 and am being a base sale for data set m), M is a