Linear regression with sparsely permuted data (Q1711600)

This paper considers the scenario of sparsely permuted data in regression analysis, in which the identifiers of the predictors (or equivalently those of the responses) are subject to an unknown permutation in such a way that a small fraction of the data is affected by a mismatch between response and predictors. It treats permuted data as outliers which motivates the use of robust regression formulations to estimate the regression parameter. The resulting estimates are subsequently used in the effort to recover the permutation. The results of two numerical experiments are presented.