A Note on NPML Estimation for Exponential Family Regression Models with Unspecified Dispersion Parameter
Jochen Einbeck, John Hinde · Durham Research Online (Durham University) · 2006
Nonparametric maximum likelihood (NPML) estimation for exponential families with unspecified dispersion parameter \\phi suffers from computational instability, which can lead to highly fluctuating EM trajectories and suboptimal solutions, in particular when \\phi is allowed to vary over mixture components. In this paper, a damped version of the EM algorithm is proposed to cope with these problems.