Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 420 - Modern Modeling Approaches for Imputation Using Survey Data
Type: Invited
Date/Time: Thursday, August 12, 2021 : 4:00 PM to 5:50 PM
Sponsor: Survey Research Methods Section
Abstract #316718
Title: Imputation Procedures in Surveys Using Nonparametric and Machine Learning Methods: An Empirical Comparison
Author(s): David Haziza* and Mehdi Dagdoug and Camelia Goga
Companies: University of Ottawa and Université de Bourgogne Franche Comté and Université de Bourgogne Franche Comté
Keywords: Additive models; Boosting; Random forests; Neareast-neighbour imputation; Cubist; BART
Abstract:

Nonparametric and machine learning methods are flexible methods for obtaining accurate predictions. Nowadays, data sets with a large number of predictors and complex structures are fairly common. In the presence of item nonresponse, nonparametric and machine learning procedures may thus provide a useful alternative to traditional imputation procedures for deriving a set of imputed values. In this presentation, we will present the results of an extensive empirical investigation that compares a number of imputation procedures in terms of bias and efficiency in a wide variety of settings.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program