Multiple Imputation Procedures Using the Gabrieleigen Algorithm

Authors: Marisol García-Peña, Sergio Arciniegas-Alarcón, Wojtek Krzanowski, Décio Barbin

GabrielEigen is a simple deterministic imputation system without structural or distributional assumptions, which uses a mixture of regression and lower-rank approximation of a matrix based on its singular value decomposition. We provide multiple imputation alternatives (MI) based on this system, by adding random quantities and generating approximate confidence intervals with different widths to the imputations using cross-validation (CV). These methods are assessed by a simulation study using real data matrices in which values are deleted randomly at different rates, and also in a case where the missing observations have a systematic pattern. The quality of the imputations is evaluated by combining the variance between imputations (Vb) and their mean squared deviations from the deleted values (B) into an overall measure (Tacc). It is shown that the best performance occurs when the interval width matches the imputation error associated with GabrielEigen.

Comments: 15 Pages.

Download: PDF

Submission history

[v1] 2016-10-01 17:00:34

Unique-IP document downloads: 32 times is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. will not be responsible for any consequences of actions that result from any form of use of any documents on this website.

Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.

comments powered by Disqus