|
Organizers |
A new algorithm for fitting generalized linear model
by
Dongwen Luo
The Institute of Information Sciences & Technology, Massey University, NZ
Coauthors: Graham Wood, Geoff Jones
The familiar geometry of linear models is extended to generalized linear models. Observations are viewed as a vector in a finite dimensional space, and the space is split into two orthogonal parts, a sufficiency space and an auxiliary space. The model surface straddles these spaces and their properties during maximum likelihood fitting are drawn out. A new algorithm for finding the maximum likelihood estimator for a generalized linear model is described using the geometry.
Date received: September 4, 2001
Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cahg-94.