Skip to main content

Statistical classification with missing covariates

Buy Article:

$51.00 plus tax (Refund Policy)

Abstract:

Summary. 

Some results related to statistical classification in the presence of missing covariates are presented. We derive representations for the best (Bayes) classifier when some of the covariates can be missing; this is done without imposing any assumptions on the underlying missing probability mechanism. Furthermore, without assuming any missingness-at-random type of conditions, we also construct Bayes consistent classifiers that do not require any imputation-based techniques. Both parametric and non-parametric situations are considered but the emphasis is on the latter. In addition to simple missingness patterns, we also consider the full Swiss cheese model, where the missing covariates can be anywhere. Both mechanics and the theoretical validity of our results are discussed.

Keywords: Bayes error; Classification; Consistency; Missing covariates

Document Type: Research Article

DOI: https://doi.org/10.1111/j.1467-9868.2007.00613.x

Affiliations: Carleton University, Ottawa, Canada

Publication date: 2007-11-01

  • Access Key
  • Free content
  • Partial Free content
  • New content
  • Open access content
  • Partial Open access content
  • Subscribed content
  • Partial Subscribed content
  • Free trial content
Cookie Policy
X
Cookie Policy
Ingenta Connect website makes use of cookies so as to keep track of data that you have filled in. I am Happy with this Find out more