Representations of efficient score for coarse data problems based on Neumann series expansion
- 72 Downloads
We derive new representations of the efficient score for the coarse data problems based on Neumann series expansion. The representations can be applied to both ignorable and nonignorable coarse data. Approximations to the new representations may be used for computing locally efficient scores in such problems. We show that many of the successive approximation approaches to the computation of the locally efficient score proposed in the literature for such problems can be derived as special cases of the representations. In addition, the representations lead to new algorithms for computing the locally efficient scores for the coarse data problems.
KeywordsAuxiliary variable Adjoint operator Coarsening at random Nonparametric information operator Projection
Unable to display preview. Download preview PDF.
- Chen, H. Y. (2009). Estimation and inference based on Neumann series approximation to locally efficient score in missing data problems. Scadinavian Journal of Statistics, in press.Google Scholar
- Robins J.M., Wang N. (1998) Discussion on the papers by Forster and Smith and Clayton et al. Journal of Royal Statistical Society, Ser. B 60: 91–93Google Scholar