Generalizing the Iterative Proportional Fitting Procedure.
Abstract
The Iterative Proportional Fitting Procedure (IPFP) can be viewed as a method for maximizing the likelihood for certain loglinear models or equivalently for minimizing the Kullback-Leibler Information between two probability densities. Both of these viewpoints lead to natural generalizations of the classical IPFP. We examine the generalizations and, with the aid of the theory, explore a practical example of expanding a contingency table.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1980
- Accession Number
- ADA084429
Entities
People
- Michael M. Meyer
Organizations
- University of Minnesota