Incremental Algorithms for Missing Data Imputation based on Recursive Partitioning
Claudio Conversano, (University of Cassino), firstname.lastname@example.org
In the framework of missing data imputation, we consider a non-parametric approach based on Information Retrieval. In particular, an incremental procedure based on the iterative use of recursive partitioning methods and a suitable Incremental Imputation Algorithm is proposed. The key idea is to define a lexicographic ordering of cases and variables so that conditional mean imputation via binary trees can be performed incrementally. A simulation study and real world applications are shown to describe the advantages and the good performance with respect to standard approaches for non-linear structures. Some possible extensions of the proposed approach to the problem of data validation will be also discussed.