Application Notes for Intel® oneAPI Math Kernel Library Summary Statistics

ID 772991
Date 12/04/2020

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

Handling Missing Values in Matrices of Observations

Summary Statistics provides the Multiple Imputation (MI) method VSL_SS_METHOD_MI to deal with missing values in a dataset. A typical usage flow is as follows:

  1. In the MI paradigm, replace each missing value with a set of m values predicted from the underlying distribution.

  2. After MI application, analyze each of the m complete datasets producing estimates and standard errors.

  3. Combine the results of the first two steps according to the rules in [Rubin1987] to produce overall estimates and standard errors.

MI approach is integrated into the library as described in [Schafer1997].