Fast Penalized Regression and Cross Validation for Tall Data with the <b>oem</b> Package

A large body of research has focused on theory and computation for variable selection techniques high dimensional data. There been substantially less work in the big "tall" data paradigm, where number variables may be large, but observations is much larger. The orthogonalizing expectation maximization (OEM) algorithm one approach penalized models which excels tall regime. oem package an efficient implementation OEM provides a multitude routines with focus data, such as function out-of-memory computation, large-scale parallel regression models. Furthermore, this paper we propose specialized cross validation, dramatically reducing computing time validation over naive implementation.