Limit points of the iterative scaling procedure
2014 (English)In: Annals of Operations Research, ISSN 0254-5330, E-ISSN 1572-9338, Vol. 215, no 1, 15-23 p.Article in journal (Refereed) Published
The iterative scaling procedure (ISP) is an algorithm which computes a sequence of matrices, starting from some given matrix. The objective is to find a matrix 'proportional' to the given matrix, having given row and column sums. In many cases, for example if the initial matrix is strictly positive, the sequence is convergent. It is known that the sequence has at most two limit points. When these are distinct, convergence to these two points can be slow. We give an efficient algorithm which finds the limit points, invoking the ISP only on subproblems for which the procedure is convergent.
Place, publisher, year, edition, pages
2014. Vol. 215, no 1, 15-23 p.
Iterative scaling procedure, Alternating divergence minimization, Biproportional fitting
IdentifiersURN: urn:nbn:se:kth:diva-144349DOI: 10.1007/s10479-013-1416-2ISI: 000332831200002ScopusID: 2-s2.0-84897629453OAI: oai:DiVA.org:kth-144349DiVA: diva2:713458
QC 201404232014-04-232014-04-222014-04-23Bibliographically approved