date: 1991
Abstract: In this paper we study Generalised Linear Models (GLM) under linear constraints on the parameters. An algorithm to estimate the parameters of a GLM under constraints is presented, based on Davidon’s algorithm in conjunction with Fisher’s scoring method, which appear to be the most appropriate methods to solve this problem. It is also suggested a starting point for the iterative search of the maximum likelihood, which performs very well when a canonical link is chosen.
Keywords: Generalised Linear Models, constraints, maximum likelihood.